Chapter 3

Explore

3.1 Welcome to Explore

Welcome to Explore. You’ve pitched your idea and it has been approved for further exploration. The main goal of Explore is to obtain evidence that your customers have a real need or problem. You will now get out of the office and speak to customers to test your assumptions. At this stage you do not yet need a large team. All you need is a small team with sufficient knowledge around the market and customers (e.g. a product manager, a UX (user experience) expert or designer, with some support from sales). The Explore stage can be split into the following phases:

1. Validate problem: This phase focuses on customer needs and jobs to be done. Do customers have the need? How are customers currently solving their problems? Are the problems currently serious enough for them to pay for a solution?
2. Value proposition: After validating customer needs, we can start testing our ideas for a solution. We can begin by creating experiments that present potential solutions to customers (e.g. landing pages). At this stage, we are not creating our product yet, we are simply testing our ideas for solutions and value propositions.
3. Market opportunity: When we validate customer needs, we will know the type of customer and market we want to pursue. As such, during Explore we will do more detailed research on the size of the total available market and whether that market is financially viable.
4. Prepare for validate: At the end of Explore, you will have confirmed customer needs and also tested some ideas for a solution. In Validate, you will begin working out how you can create a solution to sell and test the rest of your business model. So you need to define the budget for building a minimum solution and its various iterations, market it to customers and support them through their experience.

3.2 Running experiments

A key part of lean thinking is the use of evidence to make decisions. For many, innovation and product development are about the pursuit of a vision. Given the failure rates of startups and new product launches, the idea of vision alone leading to success seems more of an illusion than anything else. There is nothing wrong with chasing big audacious goals. However, our passion for our ideas must be tempered with a keen interest in reality (i.e. testing our assumptions).

To test our assumptions, we need to adopt the scientific method and apply it to innovation. The idea of science conjures up images of men and women in white coats working in a lab. However, science is merely a set of tools and methods. When we test our ideas, we are not merely collecting evidence and data. We are trying to use the evidence as a basis of learning and decision making.

Science is a hypothesis driven methodology. We cannot use the data we collect to learn and make decisions, unless we first make explicit our expectations. As such, we must turn the assumptions we identified during Idea into testable hypotheses. The best way to create hypotheses is to set minimum success criteria for our experiments. These criteria benchmark what would have to happen for us to reach the conclusion that the evidence supports our assumptions. With clear hypotheses in place, any data we collect can then be benchmarked against our criteria. This makes it easy for us to learn, make decisions and track progress.

It is also important to note that minimum success criteria cannot be developed in a vacuum. Such criteria are connected to a specific test or experiment. So, the first thing we have to do for each assumption we have is brainstorm some experiments. After we have chosen a specific experiment to run, then we can set the minimum fail criteria for that experiment. Only after this process are we now ready to get out of the building and start testing our assumptions.

3.3 Tracking progress and managing our experiments

When we run our experiments, we need a tool to track the progress of the work we are doing. Tracking progress happens at the level of a single experiment we are running, and also across experiments to see how well we are doing at reducing the number of assumptions in our value proposition or business model. For the Lean PLC, we have adopted the Experiment Canvas.¹⁶

After we have mapped out the experiment, we are now ready to get out of the building and test our assumptions. After we have finished running the experiment, we then have to analyse the data we have collected and make decisions.

Figure 3.1

3.4 Earlyvangelists¹⁷

An important question for running experiments concerns who should be recruited as participants. There are also related questions about sample size and how we know when to stop running an experiment. First, experiments during the Explore stage should target early adopters or earlyvangelists. Working with earlyvangelists¹⁸ has the potential to provide the greatest amount of learning because:

• they have the problem or need we are trying to solve
• they are aware that they have the problem or need
• they have been actively looking for a solution
• They have tried to put together a solution
• they have a budget to pay for a solution.

These characteristics make earlyvangelists, if we can find them, the best people to test our assumptions early in our process. Earlyvangelists are so keen for solution that they would be happy to give us feedback on early versions of our product. It is also important to note that if we find a customer segment in which any one of these five characteristics is missing, then it will be difficult for us to test customer needs and validate early versions of our solution. For example, if we find that customers have a need but they are not aware of it, we now have a marketing job on our hands, which is the opposite of the discovery we want to do during Explore.

Figure 3.2

With that in mind, we can turn to the question of how large our sample of participants should be. We do not have any specific guidance on this. It really depends on the type of research you are doing (e.g. B2C vs B2B). Our sense is that experiments with earlyvangelists should provide strong signals. As such, a small sample of 10–15 participants should be sufficient. If you find that you are getting mixed signals, you may need to define more stricter criteria for selecting your earlyvangelists. In contrast, if you are getting a very strong and consistent signal from customers, you can stop before you reach your target sample size.

3.5 Pivot, persevere, stop

We run experiments to be able to make decisions about what to do next. We have the choice to persevere with our idea if the evidence supports our assumptions. However, sometime the data suggest that our assumptions are wrong. In that case, we don’t want to just give up on our idea. We also don’t want to keep going by simply ignoring the evidence. We can make the decision to pivot. This is when we keep one foot in what we have learned, and we make some changes to our assumptions before we test them again. In The Lean Startup, Eric Ries¹⁹ described ten types of pivots that teams can make if they find that their assumptions are not supported by the evidence.

Figure 3.3

These are some of the options that we have when we are faced with a situation where some of our assumptions are not supported by the data. We are often asked how many pivots a team should make before they stop. This is a judgement call that depends on the strength of the signal you are getting from the market and the resources you have at hand. If the resources are available you don’t want to stop too soon. However, if you are getting really strong signals that your idea will not work, then don’t waste your resources. Stop and move on to another project!

3.6 Desk research

Desk research, also known as secondary research, is a good starting point for testing our assumptions before we get out of building. The focus here is on information that already exists. Although the data may not be specific to our hypotheses, they are great for learning about the market and our customers before we meet them face-to-face. In this age of ubiquitous information, it is amazing what data are readily available for us to review and make informed decisions.

Desk research is particularly useful for testing our assumptions about the structure and size of the available market. It is also useful for discovering any trends in population, economics and technology that may impact our product or service. We can use desk research to get information on demographic trends, suppliers, distributors, competitors, brands and products. All of these data can give us a good sense of the market environment we are about to enter.

3.7 Customer observation

Genchi Genbutsu is a Japanese phrase that means ‘actual place, actual thing’. It has been popularised as a principle of Toyota’s Production System.²⁰ The principle is based on the notion that in order to understand something you have to go to the actual place and see the actual thing you are interested in. This is why the phrase is often translated into English as ‘go and see for yourself’. As some point, we have to get up from behind our desks and see how our customers live and work in their context. Customer observation allows us to gain empathy into what our customers struggle with. Such insights can then be used to inform product development.

3.8 Customer interviews

We already highlighted the limits of simply observing our customers without talking to them. While we can learn a lot from watching people, it is often difficult to understand why people are making specific choices or taking specific actions. Furthermore, it is sometimes difficult to get access to observing people in their context. There may be constraints in terms of time, availability or even legal issues. In such circumstances, customer interviews are a quick and cheap way to gain access to customers and learn about their needs.

3.9 Empathy map

Now that you have been out of the building observing and talking to customer, it is time to create a picture of what we have learned. Furthermore, there could be members of our team who were not out of the building with us, who also need to develop an understanding of what we learned. A good tool for developing this picture is the empathy map. It allows teams to map what they saw customer say, do, feel and think.²²

Figure 3.4

3.10 Update the circle

At this point, we need to revisit the customer profile on our value proposition canvas. Remember that all the work we have been doing so far exploring customer needs was triggered by the risky assumptions we identified when we reviewed our value proposition canvas. So, the question we need answer is whether all the lessons we are learning support our assumptions and hypotheses.

• Customer jobs, pains and gains: As team, we need to review each sticky note on our value proposition canvas. For each sticky note, we need to discuss whether what we learned from customers supports that assumption. Were we correct in making that assumption about customer jobs, pains or gains? Do we need to revise the assumption? Or should we remove it from the canvas altogether?
• New learnings: We will also have learned new things about our customers that were not part of our original canvas. These new learnings will need to be added to the canvas as well. Remember to put only one job, pain or gain per sticky note.
• Make decisions: The more we get out of the building, the more our value proposition canvas will change. At some point, we will have to make decisions. Do we know enough about customers to start thinking about solutions? Have we found meaningful customer jobs to be done? How strong are the customer pains/gains? How are they currently solving their problems or meeting their needs? Do we think customer may pay for a solution to help with their jobs to be done?

The answers to these questions will inform the decision to stop, pivot or persevere. But once we have a sense that there is a real need that we can serve, it is time to revisit the value map on our value proposition canvas.

3.11 Update the square

As our customer profile has been evolving based on customer learnings, we now need to revisit our value map. It is highly like that we will have to update our value map based on learnings. Remember that our goal is to get to problem-solution fit. This means that we need the features of our product or service to match our customer jobs to be done, pain and gains.

• Product and services, pains relievers and gain creators. Given what we now know about customer jobs, pains and gains, do we still have the right product or service in mind? Do we have the right feature set to relieve the pains or create the gains that customers want?
• Learn then confirm: Innovators have a tendency to take customers at their word. This is partly because of the confirmation bias we spoke about earlier. It is important to recognise that, even if our revised value map is based on customer learnings, our ideas about solutions are still untested. So, an important principle for all innovators to follow is to learn then confirm. In other words, talk to customers then get them to do something.
• Identify assumptions: So, after we finish revising our value map, we must resist the urge to make the leap to building a product. Instead, we have to review the map for risky assumptions. Are we confident that customers would want the solution in the way we are imagining it? Will they be willing to sign up for updates, pre-order, engage in a free-trial or make a purchase?
• Early solution testing: Believe it or not, these questions can be answered without building a product. Furthermore, the questions can be answered, not with interviews, but by running experiments that get customers to do something. We will now turn to a few examples of experiments we can run. Such early solution testing, ensures that we enter the Validate stage of the Lean PLC with clearer ideas about the solution we should be building.

3.12 The landing page

The landing page is a single web page that is used to test our value proposition. You can also use flyers and posters to run similar tests. The goal of the test is to check whether our value proposition resonates with customers. Our landing page simply says what the product does and then asks customers to perform some sort of action that registers their interest.

Figure 3.5

3.13 Comprehension test

The results obtained from a landing page or a test advertising campaign are only meaningful to the extent that people understand the headline or the value proposition. It is possible for an experiment to fail simply because people did not understand the value that was offer. As such, before setting up your landing page or running a test adds, it makes sense to run a comprehension test.

3.14 The explainer video

Another way to ensure that people understand the value proposition is to create an explainer video. This is especially useful when you are thinking of creating a complex product or a product that serves complex needs. A popular example of an explainer video is the one Dropbox created before they had finished building their product. This video illustrated how the product ‘worked’ and resulted in over 50,000 signups for a product that was yet to be completed!

Figure 3.6

3.15 Paper prototypes and storyboards

If making a video is too costly an option for you, there is another way to test your value proposition visually. With paper prototypes, you can talk people through the potential experience of using your product. You can then invite them to give you feedback about your idea. You can even allow customers to scribble on the paper prototype. They can then show you the places they may look for a feature and click to find value. This feedback is powerful as an early indicator that you are headed in the right direction.

Storyboards are also a great visual way to test your value proposition. You can sketch out a narrative storyboard or a short cartoon that tells the story of how your customers can use your product to solve their problems. Such a visual presentation can help in two ways: it can help you check whether your understanding of the customer context and needs is correct; it can also help you test whether your proposed solution resonates with customers.

As shown in the image, a storyboard needs to have only a few panels that show:

Figure 3.7

• the problem you are trying to solve for customers
• the context in which customers will encounter and use your solution
• how the solution will help them solve their problem or get their jobs done
• the key benefits they will get from using your solution.

3.16 Customer journey mapping

A variation of the storyboard is the customer journey map. This map is a graphical representation of the customer’s experience via key touch points with your organisation and the solution. This process is often used for products that are already in the market. However, it can also be used for new ideas as a way to test our thinking about how our solution is going work within a customer journey.

Figure 3.8

3.17 Picnic in the graveyard

When to use it

This method is called picnic in the graveyard, because it focuses on examining products that have failed and ‘died’. The goal of all the methods and tools we have described so far is to help us get an idea of the product we should be building. What features should it have? How can we deliver value? What is the minimum set of features that can be in our first version of the product? However, even if our product idea is highly creative and original, there is nothing new under sun. It is likely that someone else has had the idea before. The picnic in the graveyard²³ method is based on finding products similar to our idea that been launched and failed in the market.²⁴

Figure 3.9

How to do it

• Explore – Take time to look for products that have been launched and failed. Desk research is good for this, but you can also speak with colleagues and customers. In fact, you can explore some of the solutions that customers are currently using that are not delivering value.
• Research – Talk to customers to learn about what they disliked about the failed product. You can also reach out to the companies or founders that launched the failed product and learn from them. Some founders may not be willing to talk, but you will find that most founders are willing to help.
• Analyse – Review your findings for hints on how you could make the products better. Many products that fail are usually close to the right solution but are missing one or two features.

  Tip

If you are going to be introducing a product that is similar to those that have gone before, then you need to be sure about your key differentiator.

3.18 The market opportunity

As the picture of your target customer and their needs becomes clearer, it is time face up to a very important question: does your target customer represent a market that is large enough to build a sustainable business? This is an important question because finding a group of customers with a need is not the same as finding a profitable market. As such, you can use the desk research techniques we described earlier to perform some market analysis.

3.19 On surveys and focus groups

So now it’s time to address the elephant in the room. You may have noticed that our review of methods left out surveys and focus groups. Although these two methods are quite popular among market researchers, there are not top of the list for us in the Lean PLC. Our concern is with getting a deep understanding of customers and their needs. We feel that these two methods are not best suited for doing that.

Surveys do have their place in the research methods toolbox, but only if they are used with other methods. The reason we don’t encourage innovators to use online or mail survey is because violates the ‘actual place actual thing’ rule (i.e. Genchi Genbutsu). When trying to learn about customer there is nothing better than going to see their lives for yourself. Surveys may be useful when you have a product out in the market. But even then, these market data must be supported by going to the market and speaking to real people.

Focus groups are also a popular market research tool. However, this method of engaging with customers increases the social desirability challenges we highlighted earlier. People’s behaviour is strongly impacted by the presence of others. It’s always important to remember that we are not just interested in hearing customers speak. We are interested in what they really think and feel. So, it is important to put customers in situations where they will be comfortable sharing their inner thoughts with us. Focus groups with strangers are not really the place for that.

3.20 Making decisions

We have reached the appropriate point in our journey to remind ourselves why we are doing all this work in the first place. During Explore, the methods and tools we have described are used to test some key assumptions: is there a real customer need, how strong is that need, what are the customer jobs to be done and what might be good solutions to meet those customer needs? At the end of every experiment, we have to review our learnings and make decisions.

Remember that if we are making something that nobody wants, then it doesn’t matter if we are on time and on budget. – Eric Ries

Our main goal as we move through the Lean PLC is to make products people want, and to figure out a profitable business model with which to deliver that value. While we do our work, we will run down rabbit holes and dead ends. As we pivot and iterate our way through Explore, we will eventually come to the end of our runway. Then it’s time to face reality:

Have we found a real customer need and potential solutions for that need?

If the answer to this question is ‘YES’, then we are ready to start building a solution. However, if the answer is ‘NO’, then we have to make a decision. Do we pivot or do we stop? This decision is a judgement call based on learnings, context and available resources. It is possible that while you have been out of the building, you have picked up a strong signal of a potentially strong need or new customer segment to pursue. In this case, if you have the funding and resources you can remain in Explore and make a pivot.

On the other hand, it is sometimes a good idea to stop. This is particularly important, if there is no real customer need. Rather than waste time and money insisting on pursuing a dead end, it is better to save our resources and deploy them in new ideas. We may not throw away our idea. We may just park it for a while. It is possible that overtime, new insights will emerge within our company that can revive our idea in some form. However, for the time being we have to STOP.

Figure 3.10

3.21 The business model

If we have found customers with a real need and we have an idea of the solution we might want to build for them, then it is time for us to start thinking about the business model. Beyond ideation about products and services, business model design allows us to take a more holistic approach to innovation. We can see how our ideas about products and services work in concert with the key partners we might need to deliver value, the channels or customer relationships we might need, potential cost structures and revenue models.

One of our favourite tools for business model design is the business model canvas. Developed by Alexander Osterwalder and colleagues,²⁵ it has nine key elements that make up typical business models. Each of these elements is described below.

Figure 3.11

• Key partners: It is sometimes important for companies to form partnerships with other organisations. This can help your company reduce risks and acquire capabilities it does not currently have.
• Key activities: In order to make the business model work, there are certain activities your team will have to do well. These can include product, problem solving, marketing or developing a network/platform.
• Key resources: This refers to the physical, human, intellectual and financial resources you will need to create value and make your business model work.
• Value propositions: This describes the value that you provide your chosen customer segments. This is a summary of the detailed information you have already captured in the Value Proposition Canvas.
• Customer relationships: This is about the type of relationship you are going to have with you customer segment. There are various types of relationships including dedicated personal assistance, self-service and automated service.
• Channels: This is how you are going to reach your customers. How will they learn about the value you offer? How will you communicate with them? And how will they get the product or service delivered to them?
• Customer segments: These are the various groups that you have identified during Explore that will be served by your product or service.
• Cost structure: This describes all the fixed and variable costs you will need to create and deliver value. Identifying risks around costs is an important part of deciding whether you have a profitable business model.
• Revenue streams: This describes how your organisation is going to capture value back from the customers. There are various revenue models you can have including asset sale, usage fees, subscription fees, renting and licensing.

3.22 Business model prototyping

The business model canvas can be used as a prototyping tool. This business model design method allows us to think through various types of models for our product or service. As we start to settle on the right solution for our customers, we need to imagine several business models we can use to create and deliver value. Prototyping prevents us from getting trapped in the local maxima of using the first business model we can think of. Below is a facilitator’s guide on how to run a business model prototyping session.

3.23 Identify assumptions – make plans to test

At this point we have achieved three key milestones. We have discovered a real customer need that we can help solve, we have identified a value proposition and potential solution that resonates with customers and we have designed our ‘Plan A’ business model. Indeed, the lessons we have learned so far will have informed our business model design. However, all this does not mean that we are ready to build and launch our product. There are still many assumptions about the solution and business model that need to be tested.

Our first business model is referred to as our Plan A for reason. We will have to iterate our way to Plan B, C, D… Z using lean innovation methods. So, as we did with our value proposition canvas, we need to identify our risky assumptions (see Figure 2.2).

• As such, the team must step back after mapping their business model and review each sticky note using the same key question: Do we have evidence for this or is it an assumption? When in doubt, it is an assumption and we must put a large red X on all sticky notes with untested assumptions.
• After this, we will then rate each untested assumption in terms of how critical it is to the success of the business model (1 = Not critical and 10 = Critical). Remember that an assumption is critical if you have to stop the project if it turns out you are wrong.
• After identifying our riskiest assumptions, we can then make plans to test to them. The next stage of the Lean PLC is the Validate stage. During this stage you will test your assumption about the solution as you are creating it. You will also test your assumptions about the business model. Your team can use the experiment canvas to start coming up with different ideas for the experiments you might run to test your ideas (see Figure 3.1). This is the critical final step as you prepare for the Validate stage.

3.24 Preparing for Validate

If you have done the activities we described above, your team is ready to ask for a larger investment to move to the Validate stage. Indeed, the product council is happy to invest a larger amount of money in your project because you have validated real customer needs and a resonant value proposition. However, Lean PLC rules still apply; do the right things at the right time. At this stage, teams should not be asking for massive investments to launch their product at scale. Instead, they should be asking for the minimal investment they need to test their assumptions around the solution and business model. We may include:

• brief details about the solution we plan to work on and how the solution supports the strategic goals of our company
• an update of the work we have done so far and how the evidence supports our proposed solution
• an update of our research on the market opportunity and whether it represents a real market
• our assumptions about the right product or service to create for customers
• our assumptions about the right business model to create and deliver value
• the assumptions we plan to test during Validate, and brief details of the tests we plan to run
• the resources and funding we think we will need to complete the iterative creative of the solution and validate our business model.

We are still not looking for a 40-page business case with five-year projections in it. The product council or investment board is looking for evidence of a real customer need that faces a large enough market and clear next steps that take us closer to a successful launch.

An interview with:
The Innovaid team at Rabobank

Rabobank is a Dutch multinational bank headquartered in Utrecht, Netherlands. It has  over 40 000 employees and is among one of the largest financial institutions in the world. The bank is a global leader in food and agriculture financing. This interview case study was conducted with Rabobank’s Innovaid team. This team is led by Siddi Wouters, Global Head of Innovation at Rabobank. Innovaid is responsible for the education and acceleration of innovation within Rabobank.