Five lessons learned on achieving product-market fit in early-stage digital startups

Most young ventures are short on resources and depend on financing from external investors. As a founder or manager these circumstances force you to deliver results fast. While the specific definition of “results” depends on your individual situation, it will likely be any variation of “achieving product-market fit” or in other words: Demonstrating that your product satisfies a market need.

At Mapudo we build a marketplace for the B2B metals distribution industry based on the hypothesis that professionals in the industry will fundamentally shift their procurement habits from offline towards digital procurement. About a year ago and already three years in business, we have decided to shift the focus back towards establishing product-market fit. Since then, we have used a variety of tools and applied different methods. Some turned out to be effective, others didn’t. Here are my five lessons learned for early-stage digital start-ups on their way to product-market fit.

1. Understand your customer

Achieving product-market fit requires you to thoroughly understand your target customers’ needs and established processes. If you operate in a B2B environment, as we at Mapudo do, the latter is particularly relevant, as customers’ reluctance to change inefficient but functioning processes is often high. Transforming user behavior with an innovative product therefore depends on understanding a market niche very well, before rolling out a refined product to the broad masses.

In order to understand your target customers’ needs and processes, I recommend recruiting a small panel of target customers that are willing to provide you with feedback on a regular basis. This could be for example via short weekly feedback calls on their most recent product experience or longer dedicated testing sessions on new concepts and features. In return, you might have to incentivize participants (e.g. with a refund, rebate, or valuable insights).

At Mapudo, we have focused too early on scaling up our business. When we realized this, we decided to focus on developing our platform based on close collaboration with target customers in a specific market niche. In order to develop a product that improves existing processes by orders of magnitude, challenge yourself to be as close to your target customers as possible.

2. Make data-driven decisions

If you build a digital product, you start to gather data the moment your product goes live. Even if this might initially be a small amount of data, make sure to base your decisions on the little but increasingly growing data available. For digital startups this first and foremost means understanding their funnel. Leverage tools such as Google Analytics to identify which parts of your customer journey work well and which parts need improvement. Implementing a tag manager will help you to define funnel events (e.g. clicks on certain call-to-action buttons) along your customer journey without recurring support of your tech team and hence facilitate the collection of relevant data.

At Mapudo, the combination of thorough funnel analysis and close collaboration with target customers has enabled us to understand that an increase in assortment led to deteriorating conversion rates, as customers had a hard time identifying the right product and vendor in long lists of products. Based on these insights, we fundamentally rebuilt our customer journey towards a configurator that identifies the right offer based on customers’ requirements.

3. Implement A/B testing

Early in the process of building a digital product, you might feel the urge to push new features live as fast as possible without validation of their actual impact. Obviously, this might work in some cases, but it bears the risk of overlooking important contingencies in other cases. At Mapudo, we learned the hard way that A/B testing is a necessity to achieving product-market fit fast. We therefore decided to A/B test each new idea, in order to validate, if it significantly (in the actual statistical sense) improves the status quo.

The good news is: Many well-established A/B testing tools are available that you can rely on. Google Optimize for example is a free tool, which lets you implement tests with a convenient frontend editor. For more advanced users, it also offers a feature that lets your tech team implement different variants, which you can assess in the Google Optimize interface. Other well-known experimentation tools are for example Optimizely, Convert.com, or Adobe Target. Whether you use one of these tools or any other tool, I clearly recommend implementing an A/B testing tool.

While many argue that A/B testing on low-traffic websites is difficult, as tests often don’t reach significance, there are strategies to leverage A/B testing, if you have low traffic:

• Run tests on your major pages with most traffic. Even consider “buying” some traffic, if your organic traffic is not yet sufficient. You can also combine tests on similar pages.
• Test bold ideas instead of small ones. Make sure your ideas are large enough to demonstrate improvements, but small enough to be implemented fast.
• Focus on high-volume test objectives. Testing on a call-to-action button on a landing page allows you to reach significance earlier than testing on overall conversion rate or net profit.
• Play with the statistical significance. You do not necessarily need a statistical significance of 95%. Use the Sample Size Calculator by Optimizely to understand the statistics.

4. Gather qualitative insights

Today, an abundance of resources to understand your customers from a qualitative perspective exists and you should clearly leverage it. Insights generated through these resources can help to bolster your quantitative data and hence develop towards product-market fit even faster. Here are four things you can start doing tomorrow:

• Observe users via session-recording tools. Tools such as Hotjar, Clicktale, or Inspectlet allow you to see videos of the screens of your visitors while browsing your website. Hence, they help you to investigate user behavior, understand your funnel, and identify errors.
• Place a feedback button on each page. This will allow you to get meaningful feedback from users, especially when things do not work as expected. You can either link the button to your contact form or simply use tools such as Survicate, Feedback Lite, or again Hotjar.
• Run surveys, on-page and via email. Implementing different kinds of surveys can help you to understand customers even better. For example, run on-page and exit surveys (e.g. by using tools such as Qualaroo, Informizely, or Hotjar) to understand why visitors behave in a certain way or why they leave. Similarly, send out surveys to existing customers (e.g. Net Promoter Score) to understand what went well and what didn’t.
• Participate in usability testing events. These events are a great and fun way to get high-quality feedback from users that are entirely focused on providing feedback. They are regularly hosted in many cities (see Usability Testessen for an overview of these events in Germany). If there is no such event in your city, you can even consider hosting your own testing event with friends and family.

5. Continuously challenge yourself

Finally, make sure to continuously question your core assumptions and remain open for the 1% possibility. As founders sometimes cannot see the wood for the trees, it makes sense to regularly let yourself be challenged by external “advisors”, e.g. other founders, investors, or any other expert in the field. Be prepared to adapt your product or business model based on their feedback and the insights you gather through the methods and tools described above. This can oftentimes feel like moving two steps ahead and one step back. As frustrating as it is, stay optimistic and keep in mind: It will allow you to improve your product on a weekly or even daily basis and reduce the risk of spending too much time developing a product that does not satisfy a market need.

How we have transformed Mapudo’s customer journey

At Mapudo, we have leveraged these methods and tools to entirely rebuild our platform in 2018. While customers were using our marketplace, our conversion rate seemed to reach a ceiling, which was significantly too low. Based on thorough data analysis as well as feedback from target customers, we understood that the specifics of the metals distribution industry made it hard for users to identify the right product from various vendors in a typical shop system based on product lists. This resulted in a too high bounce rate in the cart, where users for example realized that the minimum order value or delivery costs were too high. We realized that we would have to identify the best offers for our customers based on their requirements. In consequence, we decided to rebuild our customer journey from massive product lists towards the first product configurator within the metals distribution industry (see screenshot below). Moving away from product lists seems counterintuitive, if you are selling goods online, especially considering that many successful e-commerce companies use exactly such product lists. Our decision to rebuild our customer journey thus inherited significant risk. The methods and tools presented above have helped us to demonstrate initial results fast and at the same time make sure that we develop our customer journey step by step towards product-market fit. Today, our customer journey resembles more that of a flight search than that of a typical e-commerce shop system. And while we are still on our way towards product-market fit, we managed to more than triple our conversion rate within less than half a year.

Obviously, there is never a guarantee to achieve product-market fit and whether you manage to do so or not depends on many factors that are outside of your influence. But focusing on (1) understanding your customer, (2) making data-driven decisions, (3) implementing A/B testing, (4) gathering qualitative insights, and (5) continuously challenging yourself will hopefully help you understand, if and how you can achieve product-market fit faster.

The article is based on my experiences at Mapudo and as a coach of different startups. Hence, it does not claim to be complete. I would also love to hear your perspectives on how to achieve product-market fit fast. So, feel free to share any thoughts, additions, and critical remarks.

Thanks to Niklas, Philipp, and Anna for reviewing an earlier version of this article.