This article is part of our Retail Reloaded series. The series offers a new vision for retail businesses through an evolved use of technology, cutting-edge innovation and digitalization practices that the leading retailers are implementing now to not only survive this crisis but emerge as winners.
Our hypothesis supports three core strategic actions that leaders must take. SAVE, ACCELERATE, and SHARE. These strategies will guide you on both how to best use technology and where you should be developing new capabilities, new skills, and, importantly, a new vision for the future.
In this article, we talk about a special technique in design thinking, used by some of the most innovative companies in the world to bring revolutionary products to the market quickly, with minimum investment, technology dead ends, and people drainage involved. We guide you through the process of discovery, show you how to set up an experiment, test the hypotheses, organize interviews, pick and engage users and, finally, which tools to use for maximum efficiency.
In our time of extreme uncertainty and rapid changes, many things that seemed impossible a few months ago have quickly become acceptable, even normal. As a result of the global healthcare crisis, we have witnessed the transformation that thousands of businesses, universities, and other organizations had to undertake in order to move their operations online in days, not weeks or years. It has shown us what’s possible and how much we can do as a society.
As the old norms crumble before our very eyes, the chance for new thinking and new models has emerged both for businesses and society. This is an exciting, if slightly scary, time. For retail businesses facing an existential crisis, the need for new models has become not just necessary but unavoidable.
In the new world of Retail Reloaded, only those businesses which can predict the new sensitivities of an emerging customer and meet them on their terms, on their platform of choice, will be ready and able to set new paradigms and confident enough to lead new markets.
Retailers are pushed into a new reality. Now, more than ever, they must be skillful in uncovering the value behind an idea, before deciding to commit to any project or investment. At the same time, there is no time to lose and inaction will prove a one-way street to failure. This article considers how to approach a new retail initiative and how best to understand its potential value.
Starting a venture, building innovative retail technology and products, penetrating new markets or even creating markets that have not existed before, are inherently risky and costly business moves. In extremely volatile conditions with a large number of unknowns, like the current times, standard market analyses and planning methodologies have proven ineffective.
Fail. Learn. Learn to fail.
Stepping into the unknown territory of a new venture requires a different set of tools than the standard experience-based planning we tend to use within the current line of business. In fact, new initiatives and ventures live on assumptions that become more precise if based on deep understanding of the customer and the predictions of future market movements.
Discovery = learning by doing
This is where Discovery-driven planning comes in handy as one of the rare cognitive tools that knows how to deal with uncertainty and chaos in a cost-effective way. Money is saved in the long run, as the initiative continues to hit the mark while avoiding expensive pivots and development time.
Because so much is not known, the initial business idea usually goes through a series of transformations and adjustments as the experimentation starts to reveal the underlying truths. New insights are incorporated into strategic planning as the initiative unveils its risks and opportunities. Armed with this new knowledge, businesses become more willing and confident to progress.
Gold lies hidden in the opaque user needs which when discovered open new markets. This is opposed to venturing out to solve a gesticulated need which is good business but ends up in a head-to-head price/quality competition and not something more exciting like market creation.
Science teaches that experimentation, not guessing, leads to a solution. It is the execution of the experiment, however, that many of us are struggling to get right. What differentiates scientific experiments from design thinking experiments is the abundance of unexpected data. It is this abundance that makes the desired outcome constantly rethink itself.
Let’s consider an example. Tom Chi, a former product manager of Google Glass, talks about rapid prototyping, which is an essential part of the learning experiment. It creates something he calls “the culture of learning”. Google Glass built 150 hardware prototypes in 10 weeks for less than $20K with 3 people, gaining hundreds of key learnings that would help them define what the V1 of product should be and how to adjust their strategy going forward. They were essentially working with assumptions. Through experimenting with the actual prototypes, they learned that fixed, preconceived product ideas won’t necessarily work.
Google’s process required extensive volumes of results to be recorded and analyzed. This achieved a critical mass of knowledge and thereby contributed towards a solution whose goal was to bring them closer to the truth about the intricate relationship their customers would have with connected glasses. It was through actions that Google informed their decisions.
Let’s explore some practical methodologies in design thinking and illustrate them with examples of successfully implemented projects. We want you to acquire the sensitivity of a psychoanalyst when empathizing with customers and their habits.
The whole point of a discovery phase in a design thinking innovation methodology is to avoid excessive spending on expensive product development that, ultimately, no one cares about because the end product doesn’t respond well to the actual customer needs.
Give me six hours to chop down a tree and I will spend the first four sharpening the axe.
― Abraham Lincoln
The discovery flow in product design is all about sharpening your axe. And with modern technologies at hand, it doesn’t need to take 4 out of 6 hours. Agile businesses rely on crafting well-thought-out hypotheses and proving or disproving them through testing and conversations with users before going into the actual building or even prototyping.
Even when the delivery phase starts, discovery never stops. Why? Because perfection is an ideal that can never be reached but must be striven for. Development is informed by confirmed hypotheses and tested prototypes, but it usually only guarantees around a month of in-detail prescribed tasks, everything after that needs to keep its hypotheses loose.
Design Thinking—sounds exciting, but why do it?
Because it seeks to discover and provide solutions to those inherently human needs through a designer’s mindset, using technology as a vehicle for testing ideas, making products, services, processes, and businesses that are more in tune with customer’s reality.
It gained popularity with the boom of technology-based products since it was able to power quick prototyping, surveying, and testing. What it enables are fast and informed decisions that put the user in the center of attention, whilst taking into account technology requirements and the economic feasibility. These components when combined create an abstract tool for problem-solving and this, in turn, spurs agile development.
The availability of development teams and the decreased cost of building or using software nowadays (especially non-cutting edge software) which relies on finished, well rounded, pleasantly UXed, off-the-shelf solutions like Azure cloud, AWS, and various systems and payment gateways, have made Discovery the #1 tool for cost effective, fast-scaling ventures that resort to SaaSing the majority of their software from the Cloud and paying as they go. And cheaply so.
Design thinking methodology helps organizations answer questions big and small, addressing a moment in time within a social context. Innovative companies love this approach for its creativity, flexibility, and tolerance to mistakes. Discovery has shown extremely useful in unpredictable situations and it can help organizations navigate the post-pandemic disruption.
How will we respond to rapid changes in the retail industry? How can we effectively support employees while simultaneously keeping our business afloat? What comes next? Technology is certain to be the solution for Retail Reloaded.
Through the creative process of discovery, we consult the reality but do so in isolated conditions. This process is commonly called an experiment.
Setting up an experiment
We all know the mantra—fail fast, measure, improve, and repeat.
Discovery is a universal process. It can be applied to any situation, organization, of any size, and it works from high-level business decisions down to the feature level (which features will go into a product, their position, and so on).
We start by hypothesizing the problem area we want to solve. The problem itself or the area in which it happens can change during or after the findings from the Discovery phase. Experimentation data will uncover whether our prediction about the problem is, in fact, true. This helps us come up with a few possible scenarios. We call each of these a hypothesis and it should be tested by a target group of users.
So, basically, the process will look like this:
- Define the Problem.
- Gather all the Data around it.
- Deduce the Hypotheses.
- Find the appropriate Target Group to test them.
Hypotheses aren’t only connected with the later phases when we are testing a built solution (PoC) with target group. They work just fine in the earlier phases, where we hypothesize on the existence of the problem and talk with users to find out whether or not it exists.
For known problems, the hypothesis testing can seem obsolete (like asking users whether the shopping chart should be in the top right) but still, it’s a good place to start since new and unexpected problems can emerge.
Let’s suppose it was a familiar problem that needed no testing, when Bing took action to make their search ecosystem more beautiful and UX friendly then Google’s. Google knew that the real problem was never the search engine’s UI. They had deliberately kept theirs relatively unsophisticated. Rather, it was the quality of returned results that mattered and this is where they put their effort in. Bing spent millions, they talked with thousands of users who would most certainly have provided valuable feedback if the whole experiment wasn’t based on the wrong hypothesis (problem). Google had the problem right from the beginning. Google wanted users to leave the search page as quickly as possible because they had found what they were looking for. Bing’s pleasant UX, on the other hand, kept their users charmed, and dwell times increased. This was a self-inflicted damage to their business model as they don’t make money by keeping people on Bing but by sending them where they want to go.
A hypothesis is an informed assumption about a problem, backed by data. It consists of
Problem => Solution + (Reason) = Result
It has the following structure:
If [cause], then [effect], because [reasons].
The initial data we have when making a prediction is not confirmed, but it’s enough to get us started with the assumption. In other words, it is our best guess. Before we dive into data, we need to take into account its source. It will be a mix of data that we have gathered from previous related experiments, our old data, industry studies, whitepapers, etc.
To actually test our hypothesis, we should spend quality time interviewing real users and identifying a preliminary set of solutions, that is to say, the features which are to be incorporated into a prototype.
In addition to talking with users, we observe them from 2 angles:
- wide – tracking their lifestyle if they let us;
- narrow – observing their behavior during a specific testing session.
Finally, to complete the picture, we should get the data from their actual usage of a prototype.
Relying solely on talking can perform poorly in regards to data quality because users may be inclined to give more positive feedback when in a personal interview.
For example, when asked whether they would pay for the product, they may say: “Yeah, sure”. But, when we put a fake paywall in front of them, which when clicked triggers a popup “Sorry we’re not selling yet but you’ve helped us learn” and measure the number of clicks, we get more realistic data about our customers’ purchase propensity. This usually differs from the data obtained through customer interviews. The example of the fake “buy” button is known as the Painted Door Experiment.
Finally, before starting an experiment, it is crucial that we define precisely the metrics of the expected outcome as well as the terms of success or failure. How do we know that the solution has hit the mark?
The set of metrics to track could be defined in the area between:
- Ongoing market trends
- Industry standards
- The numbers of close competitors
- Your ambition
- The actual natural customer behavior (from which, in theory, we can capture total value in an ideal case)
Interviews require planning and a defined execution. But once a process and methodology have been established, it will become an infinitely repeatable mechanism that can be polished over time with iterations and knowledge increments.
Start by preparing a set of questions with answers that customers can choose from, but allow plenty of space for additional verbal or written feedback. Leaving customers to invent the answers completely, however, may be counterproductive. People often don’t know what they want and when they do, it may blow an experiment out of scope. We suggest the middle way – determine solutions from which to choose but leave space for creativity and unboxed thoughts.
Think about the usability of each answer: How will this information help me solve the problem?
Generally, non-biased questions that target inner motivations in a subtle way will reveal the most. Look for emotional responses. A repeated question “why” helps with getting to the core of the problem.
Moreover, we want to ask questions that are not related to the product but, instead, to the context of a person’s life. There is no universal set of questions or a style that will work for all. An experienced interviewer can take a completely open approach and still get good results. The key is to guide the interviewee from one hypothesis to the next, while picking up on important details supported by emotional responses.
In his TED lecture on the nature and success of the first discovery experiments which have changed the way we think about product innovation and consumer choices, Malcolm Gladwell, the author of five New York Times bestsellers, a radical global thinker and philosopher, gives advice on interviewing techniques:
Connect problems to opportunities. Find patterns that emerge across multiple interviews. Look at what customers are saying in support tickets, user forums, or even on social media. Gather up supporting evidence and build a case for solving the problems you uncovered as a result of the interview.
Once you’ve built up a strong case, it will inform the work on a prototype. Needless to say that once you have the prototype, you go back to testing. But, this time, your users actually use what you’ve built.
Customers often cannot articulate the problem, but they can react to the solution. Ask unambiguous, unbiased questions in an interview or a survey, show them a wireframe or a prototype, look for emotions, opinions, original ideas, etc.
Looping back when the product or the feature is done is an important part of the relationship-building process. This is called Customer Development. It’s like soft internal sales and it helps to convert customers to brand advocates. Express appreciation for the feedback received and inform the users who participated in the interview that they can now enjoy the improvements they helped to materialize. Early adopters can become your advocates in the go-to-market strategy. Some products will go a step further and offer free or discounted use of new features or products for those who helped to shape them. Besides the obvious attraction that this approach has for the customer, it helps companies ensure that the new product or feature gets vital initial traction.
Discovery is all about finding an effective solution to a concrete problem (or discovering the problem, for that matter). Unfortunately, many teams use interviews as a mechanism simply to validate their ideas, not to discover radically new ones. When discovery is used as a rubber-stamp exercise to validate ideas to which the team is already committed, the team is pushing their solution while appeasing or putting aside the facts that are informed by real customer needs. This happens more often than it should and when it does, it’s a complete waste of everyone’s time. Watch out for biased interviews!
Innovation doesn’t come from the market. Opaque needs do. The way to catch them is to observe emotion and customer behavior during the interview. More important work lies in analyzing and prioritizing customer needs on the scale of their impact — their universality and applicability on a larger set of users.
User personas need to be conceptualized as a composite sketch of a larger set of the target market, based on validated commonalities, not assumptions.
You will instinctively know this person through the data you gather reverently. Try to include in your discovery experiment the values and attitudes which stem from the wider cultural and emotional context. These lay somewhat deeper in the user. Can you understand his or her emotions, motivations, hopes, and fears? What drives a certain type of behavior? What is missing in your product to enhance their current state?
Most of these questions can be discovered by means of Psychographics, the art of knowing your customer deeply.
Alongside unlocking their psychographics, spend time with users to understand their context when using the products, and completely absorb the flows they are going through. Observe how they experience the product beyond the rational.
Products need to naturally align with existing behaviours, not opinions. A product experience does not end within itself but within a user’s life.
In the end, all the data we gather is informative. The actual innovation comes from within. Which brings us to one of the crucial key elements of Discovery — culture. Learning, curiosity, and creativity within our own team that can be nurtured and developed.
Engaging people, picking their brains
Innovation by All
—Michael C. Bush—
Great Place to Work, the company that has been in the business of workplace culture for more than 30 years, surveying millions of people around the world and advising Fortune 100 companies on leadership and transformation, stands behind the “Innovation by All” concept. The concept has emerged from their analysis of the success of the most innovative companies with whom they work.
In our interview with Michael C. Bush, the CEO of Great Place to Work, we talked about the concept of “Innovation by All” which sources ideas and solutions from all the people in a company. The approach is quite simple: the more, the merrier! People coming from different backgrounds and fields can offer fresh perspectives that most of us never considered. It is up to leaders to give these voices a platform and encouragement to speak up.
Echoing this idea in another one of our interviews with yet another prominent leader, Maya Strellar-Migotti explains how she started an innovation platform at Ericsson, where she led teams of more than five thousand people and developed products for Ericsson’s global market that served more than two billion end-users.
In collaboration with a world leader in design thinking and innovation, IDEO, Maya’s teams discovered a way to connect their worldwide offices of more than twenty thousand people to a single thinking, idea-generating body operating on an Innovation Platform, according to the unique framework they devised. Back in the 2009, this was an award-winning discovery and, ultimately, innovation project that was used by Harvard Business School as an example of effective innovation practice in large, international teams.
In their guidebook on design thinking, IDEO explains how they build resilient innovation strategies and teams:
Our teams include people who’ve trained in applied fields such as industrial design, environmental architecture, graphic design, and engineering; as well as people from law, psychology, anthropology, and many other areas. Together, we have rallied around design thinking as a way of explaining design’s applications and utility so that others can practice it, too. Design thinking uses creative activities to foster collaboration and solve problems in human-centered ways. We adopt a “beginner’s mind,” with the intent to remain open and curious, to assume nothing, and to see ambiguity as an opportunity.
Since there are infinitive use cases for discovery in business, it follows that there is an infinite number of tools to address them. We will, therefore, venture to mention just a few.
For workplace culture and employee satisfaction testing: Emprising, the tool our company is developing together with our customer, Great Place to Work, a global authority on workforce culture and leadership, used by Fortune 100 companies. We include this tool because it can help organizations transform their mindset and to underline that any new approach when building products must start from within. Company culture is possibly the hardest thing to change. Measuring progress and employee satisfaction with new ways of doing things is as essential as measuring product success. After all, it is employees that make great products.
Managing customer data and experiences across channels is indispensable for any retailer discovering how to improve its presence in the marketplace. Check Segment, as it offers a free account for smaller retailer businesses and a free trial for everyone.
Another awesome tool for prototyping, testing, and engineering web, mobile apps, TV apps, and IoT/conversational apps is the beloved Optimizely. There’s a slightly cheaper alternative, Validately, which also offers a free trial.
Since Discovery planning is all about working with assumptions and transforming them into knowledge through testing and the influx of data, it is important to create an ecosystem that will be able to produce the following:
- Construct predictions about the activities needed to produce, sell, service, and deliver the product or service to the customer. This will result in capturing a very rough initial total cost.
- Evaluate the existing industry standards and define the ways that your venture will add value to the current market or create a new market. Which are the characteristics of your product that will enable it to conquer market share or create a completely new market for itself?
- Accuracy is not the goal. The goal is to build a model of the resources, economics, and impact of the initiative. Through testing and data input, we can reflect this new knowledge and incorporate it until we attain the desired level of precision and ensure a confident investment.
- We set out provisional milestones and validation mechanisms that can help us to gauge the development of a new initiative during the discovery phase.
- Finally, Discovery is used to approximate the market value of the product/venture and expected revenues.
When a company uses this lean and open-ended approach, major flaws in the business concept will emerge early on and create the space necessary for the discovery of stronger concepts long before important investments need to be made.
Thus, the goal of Discovery is to prevent blind leaps into the unknown which turn initiatives into black holes of resource and investment. It is, essentially, a money-saving device that enables leaders to make informed decisions in times of extreme uncertainty while preparing their teams to excel at these new Retail Reloaded strategies.
In the next article on the SAVE #RetailReloaded strategy, we explore how to manage risk. What are the evidence, milestones, and expected outcomes that we need? How can we master the mechanisms by which we can transform assumptions into knowledge and, finally, ensure flawless decision making?
If you need professional technology advice on new initiative ideas that you have, use the form below to get in touch with us. We can help you discover and build products that will accelerate your digital presence with your customers and prevent you from spending resources on the initiatives that don’t stand a chance.