Customer-centric agility: how agile transformation fuels business success

What agile transformation means

Agile transformation is the shift from traditional, plan-everything-upfront ways of working to an approach that is flexible, collaborative and centred on the customer. Work is delivered in small increments, tested with real users and improved continuously.

The approach has moved well beyond software teams. In the Project Management Institute’s 2017 Pulse of the Profession survey, 71% of organisations worldwide said they used agile approaches at least some of the time. And the evidence on results is encouraging. The Standish Group’s CHAOS Report 2015, which analysed more than 10,000 software projects, found that 39% of agile projects were successful (on time, on budget and with a satisfactory result), compared with 11% of waterfall projects.

The reason is simple. Customer preferences change quickly. Agile keeps an organisation close enough to its customers to notice, and nimble enough to respond.

Keeping product development aligned with customers

Agile aligns product development with what customers need. Frequent iteration and feedback are at its heart. By releasing working increments early and often, teams stay in a feedback loop with customers and check, again and again, that they are on the right track.

Instead of building features in isolation, agile teams refine the product based on what users do and say. Visual tools such as task boards make work in progress visible, so the team can see what is changing and adjust quickly. The result is a product that evolves with its customers, rather than one that follows a plan written before anyone had used it.

What agile changes in product development

Agile replaces long, rigid phases with short cycles, usually called sprints or iterations, at the end of which priorities are reviewed. This makes three things possible.

• Flexibility: teams can respond to new requirements, market shifts or customer feedback within weeks, not at the end of a year-long plan.
• Speed: breaking work into smaller pieces means usable features reach customers much sooner than in a traditional waterfall project.
• Learning: early releases bring early feedback, so each iteration improves on the last.

The organisational effect can be significant. McKinsey’s 2021 research on agility found that highly successful agile transformations typically delivered around 30% gains in efficiency, customer satisfaction, employee engagement and operational performance, and made organisations five to ten times faster.

Three principles that keep agile customer-focused

The 2001 Manifesto for Agile Software Development set out the values and principles behind agile. Three matter most for staying close to customers.

Iterative delivery

Our highest priority is to satisfy the customer through early and continuous delivery of valuable software.

The first principle of the Manifesto for Agile Software Development, 2001

Each iteration produces something usable, which allows fast feedback and early course correction. Progress comes in manageable steps rather than one big leap, which reduces risk.

Collaboration

The manifesto values “customer collaboration over contract negotiation”. Cross-functional teams, with developers, designers, testers and business stakeholders working together, solve problems faster and stay aligned on goals. That collaboration extends to customers themselves, who are involved throughout so that the team builds the right thing.

Customer feedback

Agile teams actively seek feedback through demos, beta releases and regular check-ins, and feed it into the next iteration. The manifesto puts it plainly: “Welcome changing requirements, even late in development.” Over time, this produces a product that fits its audience.

Other principles, such as continuous improvement and a sustainable pace, matter too. But these three do the most to keep the customer at the centre.

Why customer-centric development matters

Customer-centric development means asking, at every stage: are we solving the customer’s problem and delivering what they need? Teams that work this way catch misalignment early, before a product goes too far in the wrong direction.

Agile supports this naturally. A customer-centric team might release a small feature, gather feedback straight away and refine it in the next sprint based on how people use it. The result is a product built around real behaviour rather than assumptions. Customers notice: when people feel heard, they become more engaged, more satisfied and more loyal.

Building customer feedback into every sprint

The most practical way to become more customer-centric is to make feedback part of every iteration.

1. Set up feedback channels

Make it easy for customers to tell you what they think, early and often: short surveys, in-app feedback, interviews and beta groups. Treat feedback as a steady stream, not an annual event.

2. Review feedback every sprint

Set aside time in each sprint to review new feedback with the product owner, designers and developers, and, where possible, a customer representative. Regular review turns responding to customers into a habit rather than an afterthought.

3. Prioritise it like any other work

Add feedback to the backlog and prioritise it by the value it delivers. Turn the most important items into clear user stories with acceptance criteria, so everyone knows when a problem has been solved. If several customers ask for the same improvement, that is a strong signal.

4. Close the loop

When a change is made because of feedback, tell the people who asked for it: “Thanks to your feedback, we have added the filter you requested.” Customers love to see their suggestions taken seriously, and they will keep giving you useful feedback. Internally, check whether the change solved the problem.

Built in this way, feedback becomes an engine for continuous improvement. You catch issues sooner, build features people want and keep the product in step with your customers.

Tools that help

Many tools support agile, customer-focused work. The right choice depends on the size of the team and how much structure it needs.

• Jira (Atlassian): widely used for managing backlogs, user stories and sprints on Scrum or Kanban boards. Customer feedback forms and support tools can be connected so that requests arrive in the backlog automatically, and the boards make customer-related work visible to everyone.
• Trello (Atlassian): a simpler, highly visual board of lists and cards, well suited to smaller teams and high-level roadmaps. Some companies use a shared board as a public roadmap where customers can see, and vote on, upcoming features.
• Feedback and research tools: survey tools such as SurveyMonkey and Google Forms, research platforms such as UserTesting, and feedback and product management tools such as UserVoice and Productboard help gather, organise and prioritise what customers say. Many connect to Jira and similar tools, so the most valuable requests flow straight into the backlog.

Whichever tools you choose, the aim is the same: make sure the voice of the customer is heard throughout development, and that nothing important gets lost.

What comes next for agile and customer-centric development

AI in the agile workflow

Artificial intelligence is already part of how many teams work, from coding assistants such as GitHub Copilot to tools that summarise feedback, suggest estimates or flag risks in a sprint. Used well, AI takes on routine work and leaves teams more time for the creative, judgement-heavy problems that matter most to customers.

Deeper personalisation

Customers increasingly expect products to adapt to them. McKinsey’s 2021 research found that 71% of consumers expect companies to deliver personalised interactions, and 76% are frustrated when that does not happen. It also found that faster-growing companies drive 40% more of their revenue from personalisation than slower-growing ones. Agile is well suited to this: small, frequent releases make it practical to test and refine experiences for different groups of customers.

Continuous feedback, in real time

With continuous delivery, teams can release changes far more often. In DORA’s 2024 Accelerate State of DevOps report, the highest-performing teams deployed on demand, often several times a day. Paired with live usage data, A/B testing and feature flags, feedback becomes continuous rather than periodic: teams can see how a change performs within hours and adjust accordingly.

Agile is a journey, not a project

Agile transformation is not something you finish. It is an ongoing evolution in how an organisation works. The tools and techniques will keep changing, but the goal stays the same: deliver more value to customers, sooner.

In our experience, the organisations that succeed are the ones that keep the customer at the centre of every iteration, not only at the start. Do that, and agility stops being a process and becomes a competitive advantage.

Sources

1. Project Management Institute (2017). Pulse of the Profession 2017: Success Rates Rise. https://www.pmi.org/-/media/pmi/documents/public/pdf/learning/thought-leadership/pulse/pulse-of-the-profession-2017.pdf
2. The Standish Group (2015). CHAOS Report 2015. Summary: InfoQ, Standish Group 2015 Chaos Report. https://www.infoq.com/articles/standish-chaos-2015
3. Aghina, W., Handscomb, C., Salo, O. and Thaker, S. (2021). The impact of agility: How to shape your organization to compete. McKinsey & Company. https://www.mckinsey.com/capabilities/people-and-organization/our-insights/the-impact-of-agility-how-to-shape-your-organization-to-compete
4. Beck, K. et al. (2001). Manifesto for Agile Software Development and Principles behind the Agile Manifesto. https://agilemanifesto.org/
5. Arora, N. et al. (2021). The value of getting personalization right—or wrong—is multiplying. McKinsey & Company. https://www.mckinsey.com/capabilities/growth-marketing-and-sales/our-insights/the-value-of-getting-personalization-right-or-wrong-is-multiplying
6. DORA / Google Cloud (2024). Accelerate State of DevOps Report 2024. https://dora.dev/research/2024/dora-report/

What UX design really is

User experience (UX) design is the work of making a product easy, efficient and even enjoyable to use. It is not only about how screens look. It covers the whole journey a customer has with your product, from the first impression to the last interaction, including speed, clarity, error messages and how it feels to get something done.

In a crowded market, that journey is often what separates products that grow from products that fade. One large survey illustrates this: in PwC’s 2018 Experience is everything survey of 15,000 consumers in 12 countries, 32% said they would stop doing business with a brand they loved after just one bad experience.

How UX shapes product success

Good UX is not a nice-to-have. It affects adoption, retention and satisfaction directly.

People decide quickly. In a small 2004 study of 15 people judging health websites, 94% of the reasons participants gave for mistrusting or rejecting a site were about its design. A clear, well-made interface earns trust in seconds; a confusing one loses it just as fast.

Ease brings people back. In a 2012 Google survey of US smartphone users, 74% said they were more likely to return to a mobile-friendly site.

Friction drives people away. Clumsy navigation, errors and slow screens all erode retention, and most people do not complain before they leave. They simply stop using the product, and you see it only in your churn figures.

The pattern holds across products: a seamless experience builds satisfaction and loyalty, and a frustrating one quietly drains your user base.

Common UX mistakes that hurt products

Even well-intentioned products fail when they fall into familiar traps.

Poor navigation and layout

If users cannot find what they need, they leave. Confusing menus, inconsistent layouts and unclear labels all add effort. Every screen should make it obvious where the user is and how to get where they want to go.

Slow load times

Speed is part of the experience. Google’s 2016 research on mobile sites found that 53% of visits are likely to be abandoned if a page takes longer than three seconds to load. Users read delay as a sign that something is wrong.

Unclear calls to action

If a user does not know what to do next, or cannot find the button to do it, the design has failed. Too many competing options create hesitation. Give each screen one clear primary action, and make it stand out.

Other common mistakes include neglecting mobile, presenting walls of text and ignoring the conventions people already know. The common thread: anything that causes confusion, frustration or extra effort costs you users. The good news is that all of these are avoidable.

Best practices for a seamless experience

• Know your users. Talk to them through interviews, surveys and observation. Test early prototypes with real people before you build. Research uncovers problems your team will not see from the inside.
• Make navigation intuitive. Organise content the way users think about it, use clear labels and follow familiar conventions. Add search or filters where there is a lot to find.
• Make it fast. Compress images, streamline code and test on real devices and networks. When something does take time, show progress so the user is never left wondering.
• Make the next step obvious. Use concise, action-led labels such as “Start free trial” or “Add to basket”, give the primary button room to breathe, and use visual hierarchy when there are several options.
• Design for everyone. According to the World Health Organization, around 1.3 billion people, or one in six of the world’s population, live with a significant disability. Readable type, sufficient colour contrast, alt text for images, and full keyboard and screen-reader support open your product to far more people. It is the right thing to do, and it is good business.
• Be consistent, and give feedback. When buttons, icons and layouts behave the same way everywhere, people learn the product faster. Confirm actions, explain errors clearly and show what has changed.
• Keep iterating. Once the product is live, use analytics, feedback and usability testing to find where people get stuck or drop off, then fix those moments first. Small improvements add up.

Two products that show what is at stake

Snapchat’s 2018 redesign

In early 2018, Snapchat redesigned its app, combining friends’ stories with private chats, moving celebrity and publisher content to a separate page, and changing where familiar features lived. Many loyal users found it confusing. More than 1.2 million people signed a petition asking Snap to reverse it.

In February, Kylie Jenner asked her followers on Twitter whether anyone else had stopped opening Snapchat. The next day, Snap’s shares closed around 6% lower, wiping roughly $1.3 billion from its market value, although analysts saw the tweet as a symptom of wider concern rather than its sole cause. By May, Snap had partly rolled back the changes. The lesson: even a successful product can alienate its core users if it ignores how they use it.

Quibi: an experience out of step with its audience

Quibi launched in April 2020 with $1.75 billion in funding and a promise of premium short-form video for phones. It announced it was shutting down that October, about six months later.

Many factors contributed, but several were experience decisions. At launch, users could not take screenshots or easily share clips, which cut Quibi off from the way people discover content on social media. There was no way to watch on a television until AirPlay casting was added in late May, even though lockdown had put people at home rather than on the move. And its “Turnstyle” format, which switched between portrait and landscape as the phone turned, added complexity without solving a real user need. All the funding in the world could not make up for an experience that did not fit how its audience lived.

The business case for investing in UX

For founders and business owners, UX is not only a design concern. It is an investment with a measurable return.

Design-led companies grow faster. McKinsey’s 2018 study The business value of design tracked 300 listed companies over five years. Those in the top quartile of its Design Index achieved 32 percentage points higher revenue growth and 56 percentage points higher total returns to shareholders than their industry peers.

Retention compounds. Keeping a customer usually costs less than winning a new one, and loyal customers become more valuable over time. Research by Fred Reichheld of Bain & Company found that in financial services, a 5% increase in customer retention produces more than a 25% increase in profit. A product that is easy and pleasant to use is one of the most reliable ways to keep customers.

Smoother journeys convert better. Every point of friction removed from sign-up, checkout or onboarding is a point where fewer people give up. Gains at these moments flow straight to revenue.

Good experiences reduce costs. A product that explains itself needs fewer support calls, fewer fixes and less reassurance. Satisfied users also recommend you, which is marketing you do not pay for.

Make UX a priority from day one

UX design is not about making a product look nice. It shapes how people feel when they use it, and those feelings decide whether the product succeeds. Do not treat it as something to fix later. Build a user-centred mindset into product development from the start: research before you design, test before you build, and keep learning once you launch.

In our experience, the products that win are rarely the ones with the most features. They are the ones that make the important things effortless. Listen to your users, invest in good design and keep refining. That is how a product does more than launch: it lasts.

Sources

1. PwC (2018). Experience is everything: Here’s how to get it right (press release). https://www.pwc.com/gr/en/media-centre/assets/press-release-customer-experience-is-everything.pdf
2. Sillence, E., Briggs, P., Fishwick, L. and Harris, P. (2004). Trust and mistrust of online health sites. Proceedings of CHI 2004, ACM.
3. Google (2012). What users want most from mobile sites today, with Sterling Research and SmithGeiger. https://analytics.googleblog.com/2012/09/mobile-friendly-sites-turn-visitors.html
4. Google / DoubleClick (2016). The need for mobile speed. https://blog.google/products/admanager/the-need-for-mobile-speed/
5. World Health Organization (2022). Global report on health equity for persons with disabilities. https://www.who.int/publications/i/item/9789240063600
6. 9to5Mac (2018). Snap loses $1.3B in market value as reality TV star Kylie Jenner says she no longer uses Snapchat. https://9to5mac.com/2018/02/23/snapchat-share-price/
7. TechCrunch (1 May 2018). Report on Snapchat partly reversing its redesign. https://techcrunch.com/2018/05/01/snapchat-reredesign/
8. Digital Music News (2020). Quibi is shutting down: short-form video platform gone in 6 months. https://www.digitalmusicnews.com/2020/10/21/why-is-quibi-shutting-down/
9. Digital TV Europe (2020). Quibi now viewable on TV sets with AirPlay update. https://digitaltveurope.com/2020/05/28/quibi-now-viewable-on-tv-sets-with-airplay-update
10. MobileSyrup (2020). Quibi finally lets you take screenshots. https://mobilesyrup.com/2020/07/24/quibi-finally-lets-you-take-screenshots-but-it-doesnt-work-how-you-might-think/
11. McKinsey & Company (2018). The business value of design. https://www.mckinsey.com/capabilities/growth-marketing-and-sales/our-insights/the-business-value-of-design
12. Reichheld, F. Prescription for cutting costs. Bain & Company. https://media.bain.com/Images/BB_Prescription_cutting_costs.pdf

Storytelling at the heart of inbound marketing

Inbound marketing is about attracting and keeping customers by offering value and building relationships, rather than interrupting people with sales messages. Storytelling sits at its centre. As the old sales adage puts it, facts tell, but stories sell.

A well-told story lets your audience get to know your brand on a personal level. Instead of leading with features and prices, it leads with experiences and emotions. That makes a business feel human, and makes it easier for people to connect with who you are and what you stand for.

Think about how long a film can hold your attention, compared with a 30-second advert. The difference is story. When a brand tells a good one, people pay attention for longer, remember more and feel more loyal.

Example: A software company could say, “Our app has feature X.” Or it could tell the story of a small business owner who used the app to win back five hours a week, and spent them with her family. The second version speaks to what the customer wants. By the end, the product is not just a tool. It is what made the difference.

Why stories stick

Stories leave a deeper mark than facts on their own, and there is useful evidence for it.

At Stanford, Professor Chip Heath runs a classroom exercise that he and Dan Heath describe in their book Made to Stick (2007). Students each give a one-minute persuasive speech, most of them packed with statistics; only one in ten tells a story. About ten minutes later, the class is asked what they remember. Sixty-three per cent recall the stories. Only five per cent recall any single statistic. A customer may forget your tagline or your feature list, but they will remember the story of how your product helped someone like them.

There is also research on why stories move us. Neuroeconomist Paul Zak’s lab found that emotionally engaging stories were associated with rises in two hormones: cortisol, linked to attention, and oxytocin, linked to empathy and connection. In his studies, those changes predicted how generously people responded afterwards. The oxytocin research is still debated, and some findings have not replicated well, so treat it as an explanation in progress rather than settled science. But the direction is clear: stories engage feeling as well as thought.

That emotional connection has commercial value. Research published in Harvard Business Review in 2015 found that customers who feel fully emotionally connected to a brand are, on average, 52% more valuable than customers who are merely highly satisfied. They buy more, come back more often and recommend the brand to others.

Building your brand story: a five-step guide

Emotionally connected customers are more loyal and more valuable. So how do you build a story that creates that connection? Here is the approach we use.

1. Define your core story: mission, values and purpose

Before you tell any story to your audience, be clear about who you are. Why does the company exist, beyond making money? What do you stand for? What change do you want to make in your customers’ lives?

A start-up founded to simplify personal finance, for example, might define its mission as “helping people take control of their money”. That theme can then run through every customer story it tells. Agree these elements as a team, so every story reinforces the same central message, and so you can show clearly what makes you different from competitors.

2. Know your audience, and make them the hero

Understand your target customers deeply: their needs, ambitions and frustrations. Then build your story around a hero who represents them. The brand is not the hero; it is the guide that helps the hero succeed.

The hero can be a real customer or a composite of several. If you sell project management software, your hero might be Sarah, a marketing manager juggling several campaigns with tools that no longer keep up. What matters is that your prospects recognise themselves in her: her role, her challenges, even her personality.

3. Outline the narrative: from problem to resolution

Map the story with a simple version of the hero’s journey. Who is the hero, and what does she want? What stands in her way? What has she tried that did not work? How does your product enter the story? What changes once she uses it?

For Sarah, it might look like this. She wants to deliver every campaign on time, but she is drowning in spreadsheets and email threads. She tries harder, and nothing changes. Then she finds your software, organises the team around it and starts hitting her deadlines. Six months on, she is less stressed and her team trusts the plan again.

Writing these answers down as a short story brief keeps the narrative logical and gives it an emotional payoff.

4. Write and refine

Draft the story itself. For a brand’s core story, 200 to 300 words is usually enough to say why the company exists, who it helps and how. Keep the tone consistent with your brand, and cut anything that does not build the character, show the problem or explain the resolution.

Test the draft on a few people who match your customer profile. Do they understand it? Do they feel anything? Revise until they do. Then share the final version internally, so everyone in the business tells the same story.

5. Tell it everywhere, and keep listening

Weave the story through your website, content, social media, email and sales conversations, so every touchpoint feels like part of the same narrative. Over time, people start to associate your brand with the problem you solve and the values you stand for, and begin to retell the story themselves.

Pay attention to which parts of the story people respond to, and use what you learn to sharpen the next telling. A brand story is never finished.

Three brands that told their story well

Nike: “Just Do It” and “Dream Crazy”

Nike’s “Just Do It” line, created by the agency Wieden+Kennedy in 1988, turned a sportswear company into a brand about personal determination. Its advertising rarely leads with the shoe. It leads with the athlete, and with the idea that anyone can push past their limits.

Thirty years later, the “Dream Crazy” campaign, released in September 2018 and narrated by Colin Kaepernick, extended the same story to athletes who risk everything for what they believe in. It divided opinion, but it was unmistakably Nike. Edison Trends estimated that Nike’s online sales rose by 31% in the days after the launch, compared with 17% over the same period a year earlier, and in 2019 the campaign won the Emmy for outstanding commercial.

Dove: “Real Beauty”

In 2004, Dove, Unilever’s personal-care brand, launched its Campaign for Real Beauty with the agency Ogilvy & Mather. Instead of professional models, it featured women of different ages, sizes and backgrounds, and challenged the narrow beauty standards found in most advertising. The story was about self-esteem, not soap, and it addressed something many women felt but rarely saw acknowledged.

The campaign became one of the most talked-about in advertising history: in 2015, Advertising Age named it the top campaign of the 21st century. It also gave Dove a purpose that went well beyond its products, and that purpose has guided the brand’s communication ever since.

Airbnb: “Belong Anywhere”

In July 2014, Airbnb introduced a new mission, “Belong Anywhere”, together with a new logo and redesigned app. The story was that travel is about human connection: staying with locals, not in anonymous rooms. Its content shows hosts and guests from different cultures sharing homes and experiences.

The narrative helped solve Airbnb’s biggest problem: persuading people to trust strangers. By putting real hosts and guests at the centre, it gave Airbnb something hotels could not offer, a feeling of belonging and an authentic local experience.

What these brands have in common: Nike stands for determination, Dove for real beauty and confidence, Airbnb for belonging. Each tells a story bigger than its product, and each stays true to its values. You do not need a global budget to do the same. Honest stories about your customers, or about why you started, can set you apart at any scale.

Choosing the right channels and formats

Different channels suit different kinds of story. Choose the ones your audience already uses.

• Social media: short stories in a single image, a short video or a thread. Stories on Instagram and Facebook, which disappear after 24 hours, suit behind-the-scenes moments; short-form video on Reels, TikTok or YouTube Shorts suits quick narratives. LinkedIn works well for business-to-business customer stories.
• Blogs and articles: room for depth, such as a full customer success story or the story of how the company began. Clear subheadings help readers follow the beginning, middle and end, and well-written articles are also found through search.
• Video: the most immersive format. A two-minute brand film on your website, or customers telling their stories in their own words, can convey in moments what text takes pages to say.
• Email and newsletters: a personal channel for stories told in instalments, such as a welcome series that shares how the company started, or a regular customer spotlight. Write as you would to a person, not a database.
• Your customers and partners: encourage customers to share their own experiences, and work with creators whose values match yours. Stories told by other people are often more credible than anything you say about yourself.

Repurpose your best stories across formats. A customer story can become an article, a short video and a newsletter feature. Then track which channels respond best, and adapt.

Start small

Storytelling is not a buzzword. It is a return to what has always moved people. You do not need an epic to begin: one honest customer story, told well, is enough. Share it, watch how people respond, and build from there.

Facts tell, but stories sell. They also inspire, involve and, over time, turn customers into the people who tell your story for you.

Sources

1. Heath, C. and Heath, D. (2007). Made to Stick: Why Some Ideas Survive and Others Die. Random House.
2. Zak, P. J. (2014). Why your brain loves good storytelling. Harvard Business Review. https://hbr.org/2014/10/why-your-brain-loves-good-storytelling
3. Lin, P.-Y., Grewal, N. S., Morin, C., Johnson, W. D. and Zak, P. J. (2013). Oxytocin increases the influence of public service advertisements. PLoS ONE 8(2).
4. Nave, G., Camerer, C. and McCullough, M. (2015). Does oxytocin increase trust in humans? A critical review of research. Perspectives on Psychological Science 10(6).
5. Magids, S., Zorfas, A. and Leemon, D. (2015). The new science of customer emotions. Harvard Business Review. https://hbr.org/2015/11/the-new-science-of-customer-emotions
6. Dezeen (2014). Airbnb rebrand by DesignStudio aims for visual consistency. https://www.dezeen.com/2014/07/16/airbnb-rebrand-designstudio/
7. Willamette Week (2018). Nike sales spike after Colin Kaepernick commercial debuts (reporting Edison Trends data). https://www.wweek.com/news/business/2018/09/07/nike-sales-spike-after-made-in-portland-colin-kaepernick-commercial-debuts/

In short

Most organisations already hold more data than they use. The gap is rarely the data itself. It is the habit of asking a clear question, looking honestly at the evidence and acting on what it shows. That habit is what data-driven decision-making (DDDM) describes.

The evidence that it pays is encouraging. In a study of 179 large listed companies, economists Erik Brynjolfsson, Lorin Hitt and Heekyung Kim found that firms that based decisions on data had output and productivity 5–6% higher than their other investments and use of IT would predict. And in McKinsey’s 2013 DataMatics survey of more than 400 senior executives, companies that used customer analytics intensively were 23 times more likely to outperform competitors in acquiring new customers, and the leaders in customer analytics were almost 19 times more likely than the laggards to achieve above-average profitability.

Both findings are associations rather than proof of cause, but they point the same way: organisations that decide on evidence tend to do better.

What is data-driven decision-making?

Data-driven decision-making means using facts, metrics and analysis to guide decisions, rather than relying on intuition or past experience alone. You start from a question, gather the data that can answer it, analyse it, and let what you find shape what you do next.

It does not mean switching off judgement. Experience tells you which questions are worth asking and what the numbers might be missing. Data tells you whether your instinct holds up.

From raw data to useful insight

Raw data is rarely useful on its own. Thousands of spreadsheet rows, survey responses and web analytics reports tell you very little until they are refined.

That refining happens in three stages. You collect data from the sources that matter. You clean it: remove errors and duplicates, and agree definitions so that a “customer” means the same thing in sales as it does in finance. Then you explore it for patterns, trends and anomalies.

An insight is actionable when it is clear, relevant to a decision you face, and specific enough to act on. “Sales are down” is an observation. “Sales fell because our three best-selling lines were out of stock for a third of the quarter” is an insight you can do something about.

Five steps to a data-driven decision

To make this concrete, we will follow one example throughout: a retail store whose sales fell by 10% last quarter.

1. Define the decision

Start with the question, not the data. Being precise about the decision you need to make tells you which data is relevant and what kind of analysis you need.

Example: Rather than guessing at the cause, the store manager frames the question: why did sales fall, and what would reverse the decline this quarter?

2. Collect the relevant data

Gather the data that bears directly on the question. That may come from internal sources, such as sales, inventory, customer feedback and web analytics, and, where useful, from external ones, such as market research and industry benchmarks. Accuracy matters more than volume: a decision is only as good as the data behind it.

Example: The manager pulls six months of sales and inventory records, footfall counts, customer feedback forms and recent online reviews.

3. Analyse it for insight

Look for patterns, correlations and anomalies. Depending on the question, that might mean a pivot table, a statistical test or a dashboard. Take care to separate correlation from cause, and check whether a pattern holds across different customer groups and time periods before you rely on it.

Example: Footfall held steady, but conversion, the share of visitors who buy, fell. Inventory records show that several popular items were frequently out of stock, and the feedback forms repeat the same complaint: customers could not find what they came for. The likely cause is availability, not demand.

4. Decide and act

Translate the insight into a decision, agree what success will look like, and act. Because the decision rests on evidence, it is also easier to explain to colleagues and stakeholders.

Example: The manager introduces better stock tracking and more frequent replenishment of the best-selling lines, and lets customers know by email and social media when those items are back on the shelves.

5. Measure, learn and adjust

Treat every decision as a hypothesis. Track the indicators that show whether it is working, and be ready to change course.

Example: Over the following weeks, the manager tracks conversion, stock-outs and sales. If sales recover, the decision is confirmed. If they do not, the data points to the next question, perhaps pricing or the product range.

Practical takeaway: Start with a clear question, gather good data, analyse it honestly, act, then measure. Repeated over time, this cycle builds a culture of continuous improvement.

The right tools make it easier

Business intelligence (BI) tools such as Tableau (part of Salesforce), Microsoft Power BI and Looker (part of Google Cloud) bring data from different systems into one place and present it as dashboards, charts and reports. They let people without a technical background see what is happening, drill into a chart or filter by region in a few clicks, and spot a change, such as a dip in conversion or a spike in support requests, while there is still time to act.

Tools help, but they are not the starting point. A dashboard built before anyone has agreed which decisions it should inform tends to become wallpaper. Decide what you need to know first, then choose the tool.

Evidence in, bias out

Being data-driven is as much a culture as a toolset. “What does the evidence say?” becomes a normal question in meetings, and decisions are recorded together with the data that supported them.

It also helps counter bias. The most common is confirmation bias: the tendency to look for, and favour, evidence that confirms what we already believe. A data-driven team agrees in advance what evidence would change its mind, and looks at all of it, not only the figures that support a favourite idea.

Two cautions keep this honest. First, data has biases of its own: what you choose to measure, and what you leave out, shapes what you see. Second, data has to be trusted. If people doubt its accuracy or timeliness, they will fall back on instinct, so invest in data quality and clear definitions before you scale up the analysis.

Why it matters

Data-driven decision-making gives decisions a clear rationale, which makes them easier to explain, and easier to learn from when they go wrong. It also makes an organisation more responsive: when you monitor the right indicators, you catch problems early and spot opportunities sooner.

In our experience, the organisations that benefit most are not the ones with the most data. They are the ones that are clearest about the decisions they need to make.

Key takeaway: Data-driven decision-making turns information into insight, and insight into action. Lead with a clear question, combine evidence with experience, and keep measuring.

Sources

1. Brynjolfsson, E., Hitt, L. M. and Kim, H. H. (2011). Strength in Numbers: How Does Data-Driven Decisionmaking Affect Firm Performance? Proceedings of the International Conference on Information Systems (ICIS). https://papers.ssrn.com/abstract=1819486
2. Bokman, A., Fiedler, L., Perrey, J. and Pickersgill, A. (2014). Five facts: How customer analytics boosts corporate performance. McKinsey & Company. https://www.mckinsey.com/capabilities/growth-marketing-and-sales/our-insights/five-facts-how-customer-analytics-boosts-corporate-performance
3. Tableau. A guide to data-driven decision making. https://www.tableau.com/learn/articles/data-driven-decision-making
4. IBM. What is data-driven decision-making? https://www.ibm.com/think/topics/data-driven-decision-making
5. IBM. What is business intelligence (BI)? https://www.ibm.com/think/topics/business-intelligence
6. American Psychological Association. APA Dictionary of Psychology: confirmation bias. https://dictionary.apa.org/confirmation-bias

Where things stand

AI has moved from experiment to everyday tool. In McKinsey’s 2026 State of AI survey of 1,719 respondents, nearly nine in ten said their organisation regularly uses AI in at least one business function, and the revenue gains they reported were most often attributed to AI in marketing and sales. In Salesforce’s 2026 State of Marketing report, a survey of 4,450 marketing decision-makers, 75% said they were using AI.

But using AI is not the same as getting value from it. The teams that benefit are clear about which jobs to hand to AI, which to keep, and how to check the results.

What AI does well today

• First drafts and variations. Headlines, ad copy, product descriptions and email subject lines, produced in many versions for people to choose from and refine. Creative can be resized and adapted for each channel in minutes rather than days.
• Personalisation at scale. Tailoring emails, offers and website content to different groups of customers, something that used to need far more people and time.
• Media buying. The major ad platforms now automate much of the bidding, targeting and placement within limits you set.
• Analysis. Summarising customer feedback and reviews, spotting patterns in campaign results, and flagging what has changed.
• Routine customer questions. Chat assistants that answer simple, common questions at any hour, provided they are tightly limited and checked.

Example: Klarna says AI helped it cut sales and marketing spending by 11% in the first quarter of 2024, with AI accounting for 37% of the savings, about $10 million a year, and that image production time fell from six weeks to seven days. These are the company’s own figures, but they show where the early gains tend to come from: volume, speed and production costs.

What should stay human

AI is good at producing options. It is not good at knowing which option is right for your brand, whether a claim is true, or who is accountable when something goes wrong.

Strategy and positioning. Deciding who you serve, what you stand for and how you are different is a judgement about your market and your business. AI can help with research; it cannot make the call.

Brand voice and taste. Coca-Cola’s AI-generated Christmas adverts drew criticism in 2024 and again in 2025; in 2024, some viewers called the advert “soulless”. The technology worked; the reaction was about what the audience felt was missing. AI can produce a hundred versions of an idea. Deciding which one is true to your brand is still a human job.

Facts and claims. Generative AI can state false information with confidence. In a 2025 study by the European Broadcasting Union and the BBC, 45% of AI assistants’ answers about news had at least one significant issue. Earlier, in 2023, the technology site CNET corrected 41 of the 77 stories it had published with AI help. Every fact, figure and claim in AI-assisted marketing needs a human check before it goes out.

Accountability. In February 2024, a Canadian tribunal ordered Air Canada to compensate a customer who had been given wrong information about bereavement fares by the airline’s website chatbot. The tribunal said Air Canada was, in effect, suggesting the chatbot was “a separate legal entity that is responsible for its own actions”, and rejected the argument.

It should be obvious to Air Canada that it is responsible for all the information on its website.

Civil Resolution Tribunal of British Columbia, Moffatt v. Air Canada, 2024

The rules that already apply

You do not need to wait for new AI laws to know the basics.

Advertising rules apply however an ad was made. The UK’s Advertising Standards Authority has said its rules apply “regardless of how the ad was created”. In 2025, it confirmed there is no blanket legal requirement in the UK to disclose the use of AI in ads, but that a disclosure is “very unlikely to mitigate the harm caused by a fundamentally misleading message”, for example an AI-generated image showing the effect of a cosmetic product.

Data protection applies to what you feed in. Personal data you put into AI tools is covered by UK GDPR, so check what your tools do with it before you use customer data. The Information Commissioner’s Office publishes guidance on AI and data protection, which it is reviewing following the Data (Use and Access) Act 2025, parts of which took effect in February 2026.

Marketing into the EU brings extra duties. If your marketing reaches people in the EU, the transparency duties in Article 50 of the EU AI Act have applied since 2 August 2026. They include telling people when they are talking to an AI system and labelling deepfakes. The “digital omnibus” changes that came into force in July 2026 delayed the rules for high-risk AI systems, not these duties, apart from a short grace period for watermarking by AI systems already on the market.

Search is changing too

AI is changing how people find you, not only how you market. Google began showing AI-generated summaries, called AI Overviews, at the top of search results in the US in May 2024 and in the UK that August; by May 2025 they had reached more than 200 countries and territories. A 2025 Pew Research Center study of US users’ browsing found that people clicked a traditional search result in 8% of visits where an AI summary appeared, compared with 15% of visits where it did not.

For marketers, that means being the source an AI summary draws on: clear, specific, well-structured content that answers real questions, from a brand people trust. It also means measuring more than clicks. Meanwhile, Google has kept third-party cookies in Chrome rather than removing them, and in October 2025 it announced it would retire most of the Privacy Sandbox tools it had built to replace them. Your own customer relationships and consented first-party data remain the most dependable foundation.

Where to start

Most teams get more from a few well-chosen uses than from trying AI everywhere at once.

1. List the work

Write down the recurring marketing tasks your team does each month and how long each takes. The best candidates are frequent, time-consuming and easy to check.

2. Start with one or two uses

Pick tasks where a mistake is cheap and a human reviews the output anyway, such as first drafts, ad variations or summarising feedback. Leave customer-facing automation until you have rules and checks in place.

3. Write down the rules

Agree what data may go into which tools, who checks facts and claims, when you disclose AI use, and who signs off. A one-page policy is enough to start.

4. Measure against a baseline

Record how long the task took and how well it performed before AI, then compare. Count the time spent checking and fixing, not only the time saved.

5. Train people and review often

Tools and rules are changing fast. Give your team time to learn, share what works, and look again at your uses and your policy every few months.

The human edge

AI will keep getting better at producing content. That makes the human parts of marketing more valuable, not less: knowing your customers, deciding what your brand stands for, and taking responsibility for what you put into the world.

The teams that get the most from AI are not the ones that use it most. They are the ones that are clearest about where it fits, and where it does not.

Key takeaway: Automate the volume, keep the judgement. Use AI for drafts, variations and analysis; keep strategy, brand voice, fact-checking and accountability with people; and write down the rules before you scale.

Sources

1. McKinsey & Company (25 August 2026). The state of AI in 2026: On the road to ROI. https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai
2. Salesforce (2026). State of Marketing, tenth edition. https://www.salesforce.com/blog/tenth-state-of-marketing/
3. Payments Dive (2024). Klarna uses AI to cut marketing costs. https://www.paymentsdive.com/news/klarna-gen-ai-openai-cut-marketing-spend-efficiency/717373/
4. NBC News (2024). Report on reactions to Coca-Cola’s AI-generated holiday adverts. https://www.nbcnews.com/news/amp/rcna180665
5. Forbes (2025). Coca-Cola sparks backlash with AI-generated Christmas ad, again. https://www.forbes.com/sites/danidiplacido/2025/11/04/coca-cola-sparks-backlash-with-ai-generated-christmas-ad-again/
6. European Broadcasting Union (2025). News Integrity in AI Assistants (press release). https://www.ebu.ch/news/2025/10/ai-s-systemic-distortion-of-news-is-consistent-across-languages-and-territories-international-study-by-public-service-broadcaste
7. The Verge (2023). CNET found errors in more than half of its AI-written stories. https://www.theverge.com/2023/1/25/23571082/cnet-ai-written-stories-errors-corrections-red-ventures
8. Civil Resolution Tribunal of British Columbia (2024). Moffatt v. Air Canada, 2024 BCCRT 149. https://decisions.civilresolutionbc.ca/crt/crtd/en/525448/1/document.do
9. Advertising Standards Authority (2023). Generative AI and advertising: decoding AI regulation. https://www.asa.org.uk/news/generative-ai-advertising-decoding-ai-regulation.html
10. Advertising Standards Authority (2025). Disclosure of AI in advertising: striking the balance between creativity and responsibility. https://www.asa.org.uk/news/disclosure-of-ai-in-advertising-striking-the-balance-between-creativity-and-responsibility.html
11. Information Commissioner’s Office. Guidance on AI and data protection. https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/artificial-intelligence/guidance-on-ai-and-data-protection/
12. legislation.gov.uk (2026). The Data (Use and Access) Act 2025 (Commencement No. 6 and Transitional and Saving Provisions) Regulations 2026. https://www.legislation.gov.uk/uksi/2026/82/made
13. White & Case (2026). EU AI Omnibus enters into force, amending the AI Act. https://www.whitecase.com/insight-alert/eu-ai-omnibus-enters-force-amending-ai-act
14. Google (2024). Generative AI in Search: let Google do the searching for you. https://blog.google/products/search/generative-ai-google-search-may-2024/
15. Google (2024). New ways to connect to the web with AI Overviews. https://blog.google/products/search/new-ways-to-connect-to-the-web-with-ai-overviews/
16. Google (2025). AI Overviews are now available in over 200 countries and territories, and more than 40 languages. https://blog.google/products/search/ai-overview-expansion-may-2025-update/
17. Pew Research Center (2025). Google users are less likely to click on links when an AI summary appears in the results. https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/
18. Google Privacy Sandbox (2025). Update on plans for Privacy Sandbox technologies. https://privacysandbox.google.com/blog/update-on-plans-for-privacy-sandbox-technologies

In short

Optimising means improving what you already do: lowering costs, raising quality, selling more of what customers already buy. Innovating means doing something new: a new product, a new market or a new way of making money. Both matter. The hard part is the balance, and the right balance changes as your market changes.

Most companies lean heavily towards the first. In PwC’s 2025 Global CEO Survey of 4,701 chief executives, on average only 7% of revenue over the previous five years had come from distinct new businesses. In the same survey, 42% of CEOs said their company would not be viable beyond ten years if it continued on its current path.

Two kinds of work

In 1991, the organisational theorist James March described two kinds of activity that every organisation has to balance. Exploitation, in his words, “includes such things as refinement, choice, production, efficiency, selection, implementation, execution”. Exploration “includes things captured by terms such as search, variation, risk taking, experimentation, play, flexibility, discovery, innovation”.

The two compete for the same people, money and attention, and they pay back on different clocks. In a later paper, Daniel Levinthal and March pointed out that the returns from exploitation tend to be more certain and to arrive sooner and closer to home than the returns from exploration. That is why, left to themselves, most organisations drift towards optimising. The pull of the short term is strong: in PwC’s 2026 survey, chief executives said they spend 47% of their time on issues with a horizon of less than a year, and 16% on decisions looking more than five years ahead.

Why doing only one fails

March was clear that both extremes are dangerous.

Adaptive systems that engage in exploration to the exclusion of exploitation are likely to find that they suffer the costs of experimentation without gaining many of its benefits.

James G. March, Exploration and exploitation in organizational learning, 1991

The opposite is just as risky. Those that only exploit, he wrote, “are likely to find themselves trapped in suboptimal stable equilibria”.

Levinthal and March gave the two dangers names. In the success trap, what works today attracts more and more investment, so what might work tomorrow gets less and less. In the failure trap, an organisation moves from one new idea to the next without staying with any of them long enough to build something durable. Two companies from the same industry show the difference.

Kodak: the success trap

Kodak did not miss digital photography. One of its engineers, Steven Sasson, built what the IEEE recognises as the first handheld digital camera in 1975, and in 2005 Kodak led the US digital camera market. What it could not do was build a business around digital that was anywhere near as profitable as film. In January 2012, Kodak filed for Chapter 11 bankruptcy protection.

Fujifilm: working on both

Fujifilm faced the same collapse in film. In 2000, according to Shigetaka Komori, its long-serving chief executive, photographic products made up 60% of its sales and two-thirds of its profit. It kept and restructured its imaging business while moving its chemical and materials know-how into new markets, including healthcare and advanced materials, and in 2007 it launched a skincare line, Astalift. In the year to March 2024, healthcare alone brought in about a third of the group’s revenue.

Signals that tell you which you need now

There is no formula, but there are reliable signs.

Lean towards optimising when

• demand is steady and customers still choose you for the same reasons;
• margins are under pressure from costs you can control;
• quality, speed or service problems are costing you customers;
• you have a clear lead, and the bigger risk is wasting it.

Lean towards innovating when

• growth in your core market has slowed, and price is becoming the main way to win;
• customers are starting to solve the problem a different way;
• a new technology or regulation is changing what customers expect;
• most of your revenue depends on one product, one customer or one market.

Most businesses will recognise signals from both lists. That is the point: the question is rarely which one, but how much of each.

How to run both at once

1. Decide the split deliberately

Make the balance an explicit decision rather than the result of whoever argues hardest. In a widely cited 2012 Harvard Business Review article, Bansi Nagji and Geoff Tuff of Monitor Group reported that outperforming firms in the industrial, technology and consumer goods sectors typically allocated about 70% of their innovation resources to core offerings, 20% to adjacent efforts and 10% to transformational initiatives. The returns, they found, tended to run the other way, with about 70% coming from the transformational work.

Treat those numbers as a reference point, not a target. They come from the consultancy’s own research rather than a peer-reviewed study, and the authors themselves stress that “the ideal balance will differ from industry to industry and company to company”.

2. Give new work its own home

New ventures rarely survive inside a unit that is judged on this quarter’s efficiency. In a study of 35 attempts to launch breakthrough innovations, Charles O’Reilly and Michael Tushman found that more than 90% of the attempts that used an “ambidextrous” structure succeeded. These kept new, exploratory units separate from the established business, with tight links at senior level. Of the attempts run within the existing functional structure, only 25% succeeded. It was a small study, but the lesson is a practical one: protect new work, and connect it at the top.

3. Measure each on its own terms

Judge the core on margin, efficiency and customer retention. Judge new work first on learning: evidence that customers want it, then early revenue, then profit. Holding a six-month-old venture to the standards of a mature business is one of the quickest ways to kill it.

4. Fund new bets in stages

Start with small, cheap tests. Release more money as the evidence builds, and stop early when it does not. That way, failures stay small and successes get the support they need.

5. Revisit the balance every year

Markets move, and the right split moves with them. Set a date, at least once a year, to look again at where your effort is going and whether the signals have changed.

Start with an honest audit

Before deciding anything, answer three questions with real numbers. Where do your revenue and profit come from today, and how exposed are they? How much of your team’s time goes on improving the core, and how much on building something new? And what would have to happen for your core business to be smaller in five years’ time, and are you already seeing signs of it?

The businesses that get this right do not treat it as a one-off strategy exercise. They make the balance an explicit decision, review it regularly, and change it when the market tells them to.

Key takeaway: Optimising keeps you competitive today; innovating keeps you relevant tomorrow. Decide the balance on purpose, protect new work from the pressures of the core, and look at the split again every year.

Sources

1. March, J. G. (1991). Exploration and exploitation in organizational learning. Organization Science 2(1), 71–87. https://doi.org/10.1287/orsc.2.1.71
2. Levinthal, D. A. and March, J. G. (1993). The myopia of learning. Strategic Management Journal 14(S2), 95–112. https://doi.org/10.1002/smj.4250141009
3. PwC (2025). 28th Annual Global CEO Survey (press release). https://www.pwc.com/gx/en/news-room/press-releases/2025/pwc-2025-global-ceo-survey.html
4. PwC (2026). 29th Annual Global CEO Survey (press release). https://www.pwc.com/gx/en/news-room/press-releases/2026/pwc-2026-global-ceo-survey.html
5. IEEE (2022). Milestone: Handheld Digital Camera, 1975. Engineering and Technology History Wiki. https://ethw.org/Milestones:Handheld_Digital_Camera%2C_1975
6. NBC News / Associated Press (2006). Report on IDC figures for the 2005 US digital camera market. https://www.nbcnews.com/id/wbna11258928
7. CNN Money (2012). Kodak files for bankruptcy. https://money.cnn.com/2012/01/19/news/companies/kodak_bankruptcy/
8. Printing Impressions (2015). New book: Fujifilm CEO explains company’s dramatic transformation. https://www.piworld.com/article/new-book-fujifilm-ceo-explains-companys-dramatic-transformation/
9. CosmeticsDesign-Asia (2015). A look at Fujifilm’s astaxanthin breakthrough in skin care. https://cosmeticsdesign-asia.com/Article/2015/05/06/A-look-at-Fujifilm-s-astaxanthin-breakthrough-in-skin-care
10. Fujifilm (2024). Financial results for the fiscal year ended March 31, 2024. https://www.fujifilm.com/de/en/news/Fujifilm-Announces-Financial-Results-for-Fiscal-Year-Ended-March-31-2024
11. Nagji, B. and Tuff, G. (2012). Managing your innovation portfolio. Harvard Business Review. https://hbr.org/2012/05/managing-your-innovation-portfolio
12. O’Reilly, C. A. and Tushman, M. L. (2004). The ambidextrous organization. Harvard Business Review. https://hbr.org/2004/04/the-ambidextrous-organization