What's the best way to test your hypotheses with minimum viable products?

The process of 'Hypothesis Driven Design' is very helpful here; especially when using a consistent format throughout the process. Both in 1) clearly defining your assumptions, specified by target group segments, as for 2) defining the metrics for measuring their effect. A proven format to do so is: 1. We believe that [offering this functionality], for [target group segment] 2. Will result in [this user behaviour] 3. Which is proved by [this quantified outcome] To define valuable hypothesis, describing the assumption is key. It helps to first turn assumptions into a question (e.g. through the IDEO 'How Might We'-format). Then, writing out several solutions to that question will bring you to a clear assumption to kick-off your hypothesis.

Here's a template for creating your initial hypothesis: There are [Persona(s)] who have certain [Jobs-to-be-Done(s)]. They might be [using certain Alternatives], but if we [offer them our Value Proposition], we can expect success through [Key Metrics: acquisition/onboarding/engagement/retention/LTV, etc.]

Don’t be too vague or unclear so that you can retrospectively fit data into the “proven” bucket; rather, be as specific as possible so that you can really see if your hypothesis is accurate or not — there’s nothing worse than putting time and capital into solving a problem that doesn’t exist at scale. Be brutally honest with yourself and your stakeholders at this point — if the data comes back that you’re on to something, get excited. But if not, don’t stubbornly persist.

The relationship between hypotheses testing and MVP development is inherently symbiotic. To streamline this process, an articulate approach involving fundamental steps is required. The first step in this process is defining the hypothesis Testing your hypothesis with an MVP is a highly effective, iterative process. It ensures efficient resource utilization and minimizes the risk of failure by validating key assumptions through real-world user interactions. You objectively accept or reject your hypothesis by interpreting user feedback and behaviors, allowing you to decide your product's future course. The MVP operates on the principles of Lean startups: build, measure, learn, enabling you to make informed decisions every step of the way.

Creating an MVP starts with a clear hypothesis that outlines assumptions about customers, problems, solutions, and value. Define the target segment, value proposition, and key metrics for validation. For instance, targeting busy urban professionals with quick, healthy meal delivery, measuring sign-ups, orders, order value, and satisfaction score. This structured approach ensures focused learning and effective MVP development.

There is a misconception that MVP always means a 'software prototype' / landing page / waitlist etc. It can also mean the smallest working 'hardware prototype' / 3D printed working prototype. It is ideally, the smallest /safest / simplest working example that customers would buy or investors would invest. You can scrap a prototype and start again, but if you spend $$$$ and no one buys your product the damage is more severe.

Alternatively, an MVP should be all three things the acronym stands for: minimal, yes, but also VIABLE and a PRODUCT. A “wizard of Oz” is NOT an MVP, but rather can be a good pre-MVP way to test or disprove your hypothesis.

Another way to test assumptions is the ‚fake door test’. Test for example a feature idea by placing a CTA on your website or within an ad for it. Without even writing a single line of code, you‘ll see whether it sparks interest in your target group or not. What awaits your audience after the click is up to your imagination: be it a 404 site or a form to collect contact details from your lead, to inform them when ready. Happy testing!

PRETOTYPE AND TEST LIKE TESLA As innovation facilitator I like to share how Tesla interested buyers for their first car. In 2003, the founders of Tesla had an ambitious idea (an all-electric 2-door sports car). In order to convince potential buyers to order its car, Tesla created a pretotype of what the car would look like. Lotus supplied Tesla with a ‘glider’ Elise - which was filled with models of key components. This was not a prototype, because the vehicle didn’t function. Tesla was able to show buyers a close proxy for the final design. Tesla asked “Will you put down a $5,000 deposit to have it built to order?” From this test Tesla secured several hundred deposits, a non-trivial result to reassure their Investors.

Dont code. It is the most expensive thing to do. Each new line of code introduces fresh bugs. releases are delayed. Moreover, the 1st version always falls short of expectations. 1st step should be to engage with Founders who had the same hypothesis, the same approach as yours but failed. You should be convinced that you have a different path. 2nd step is to design experiments that allow you to test parts of your hypothesis using specific channels -such as digital advertising, social networks, or WhatsApp, without coding a single line 3rd is to explore existing tools and platforms like Wordpress, shopify. They might seem suboptimal but those optimizations are not to be chased on day 1.

Building and launching your MVP is all about agility and efficiency. Prioritize functionality over frills, aiming for swift deployment and data collection. Leverage existing platforms and tools to expedite the process. Remember, the key is to focus on your core value proposition and incorporate only crucial features. Genuine user feedback from your target segment is the golden ticket to refining and advancing your startup. This iterative approach is the heartbeat of progress.

At the risk of being redundant, if your MVP isn’t actually viable it’s not truly possible to test your hypothesis. It’s one thing for people to say they would use a theoretical product, an entirely different thing for people to actually use it once it’s been built.

In my experience, here's something that has worked Early Engagement: Involve users from the start. Seek their input on concept, design, and features. Are you solving a real problem? Usability Testing: Evaluate user-friendliness. Ask users to perform tasks. Where do they struggle or get stuck? Feedback Loops: Create channels for continuous feedback. What do users like or dislike about the MVP? Any suggestions for improvement? Performance Metrics: Define success metrics. Are users achieving their goals? Measure engagement, retention, and conversion rates. Scalability: Assess if the MVP can handle growth. Can it accommodate more users and data?

I would recommend getting to this stage as quickly as possible. Far too many times I see entrepreneurs waiting years to even build the MVP. I agree with the above points, utilise simple software to help get the product set up. Investors will want to see the concept. Users will response when you enter an alpha test mode. You must validate your idea with feedback and the only way to do this is to get your MVP built. Get it out of your head and into the world. You shall not regret it!

We must ensure on focusing on building a version of the product that includes only the essential features necessary to test your primary hypotheses. This lean approach ensures us to spend minimal resources while validating the core value proposition. Also, utilizing tools and platforms that facilitate rapid development, such as no-code platforms like Bubble, Webflow, or prototyping tools like Figma and InVision is also important.

Based on a hypothesis, combining qualitative data (e.g. web analytics) with quantitative data (e.g. user interviews) is a proven process to gain the most valuable insights. Either you indicate pain points through actual (web) analytics and other quantitative sources –which I call the *Where* ('does it go wrong')– and find out through user research observation what the users' behaviour is around that specific area (called the *How* ('does it go wrong'), or vice versa; Involving a small group of users in a scenario to observe how they behave (e.g. in a prototype) to later on focus in the web analytics on specific pain points indicated by that qualitative research. Key takeaway: don't forget to go out there and talk to your actual users!

Common Themes: Look for recurring themes in feedback and data. Iterative Improvement: Feature Adjustments: If a feature didn’t work as expected, consider adjusting or reimagining it based on user feedback. Pivot: If the data and feedback indicate a fundamental flaw in your concept, be open to pivoting your business idea. Persevere: If the core idea is validated but certain features need improvement, persevere and iterate on the existing concept. Communication and Transparency: Internal Team: Ensure everyone understands the feedback and its implications. Future Strategy: Strategic Adjustments: Adjust your overall business and marketing strategy based on the insights gained.

Analysing data points will be critical to a successful outcome and improvement of your products and services. So, set those systems up early in order to measure success. This could be as simple as integrating Google Analytics. You must identify which products and services are in demand and performing well. Use that data to improve and push those services for better revenue and customer growth. As a startup your attention to detail will give you in a competitive advantage. Spend time at the start of your journey to ensure you have the systems in place to gather and data and then pay special attention to it!

Absolutely, the journey doesn't end at MVP launch—it's just the beginning. Rigorous analysis of data and feedback is crucial. This is where you separate what works from what doesn't. Define clear success metrics and scrutinize the results for insights. These nuggets of wisdom are the fuel for refining your product and strategy. Remember, the beauty of MVPs lies in their iterative nature. They're a compass guiding you toward success, minimizing risks and maximizing learning from your valued customers.

As a startup founder, analyzing and iterating post MVP launch is crucial for my growth strategy. I would assess data collected using predefined key metrics, seeking patterns and trends to evaluate if the MVP supports my initial hypotheses. Unforeseen findings would be seen as opportunities to understand the customer and the market better. Based on analysis, I would make informed decisions on whether to pivot (alter approach), persevere (continue with current direction), or kill (stop the project). This iterative process allows for continuous learning, refining the product, and optimizing the business strategy, ensuring that I am effectively responding to real user needs.

My best advice when it comes to testing out your hypothesis is to just get started and test it! Too many founders think they have to have it all figured out to start. You don’t. Just figure out A to B, then B to C, and continue. Filling out a basic Typeform survey and sending to potential users is a great starting point to dive a bit deeper into your idea and start the validation process.

"Turning your startup vision into reality hinges on one crucial step: MVP testing." 1. Define Clear Hypotheses: Identify key assumptions about your customers, their problems, and your solution. 2. Simplicity: Your MVP should be the most basic version that allows testing of your main hypothesis. 3. Choose Metrics: Decide on measurable outcomes that will indicate success or failure. 4. Deploy to Target Audience: Offer your MVP to a subset of your target customers. 5. Collect Feedback: Use surveys, interviews, and usage data to understand user experiences and preferences. 6. Analyze & Iterate: Review feedback and metrics. Did the MVP validate your hypothesis? If not, why? Adjust and test again.

MVPs are very popular these days, but don't underestimate the potential of other approaches! The messy early phases of developing your ideas have huge potential to save you trouble down the road. In his book, The Right It, Alberto Savoia introduces the idea of a PRETOTYPE – the very, very early version of a prototype. For example, before even start writing a book, you could print a fictional cover with a potential title, wrap it around a book, and talk to people at a local bookstore. For a few bucks, you can gain loads of insight and decide if it's even worth to go the next steps!

As a startup founder, beyond the MVP, I'd also have to consider further elements like scalability, competitive landscape, potential legal issues and market trends. Another crucial factor would be the financial viability, considering both the current burn rate and future funding needs. Customer communication cannot be overlooked, ensuring that their feedback is heard and acted upon, improving customer adoption and retention. Employee morale and team dynamics, particularly during periods of pivot or major changes, would be vital. Lastly, while relying on data, it's equally important not to ignore the intuition and vision that led to the establishment of the startup in the first place.

As Steve Blank said, get out of the building. Most founders test hypothesis by doing online questionnaires. This is the biggest rookie mistake I see every year. The best way to test your hypothesis is to actually speak to people. Meet with people and do focus group sessions. Ask what alternatives they use, why they're okay to live with existing products/services, what would make them change their behaviour etc. While you have digital tools to test hypotheses, you can't accurately test them sitting behind a computer.