I thought I’d share our experience doing the Google Ventures 5 day design sprint based on the book SPRINT : https://www.thesprintbook.com/. I am going to go through our journey so that you can determine if this process is suitable for your startup!
Background
Two friends and I (tech backgrounds) decided to spend a week trying to test out a business hypothesis. I had done a design sprint at a previous startup when we were trying to pivot and thought this would be a great structure for us to come up with something concrete by the end of a week. We decided on a general problem space of helping small e-commerce stores based on the thesis that as people lost jobs and had to work from home post covid, the amount of people starting one man (person) online shops would increase. Shopify’s latest numbers are proof. My last startup worked with Twitch influencers as the main customer acquisition channel, and I knew first hand what a disaster the process of working with influencers/content creators was/is.
Problem Space
We used a modified version of the process, found here https://www.invisionapp.com/inside-design/design-sprint-2/ and started with the Expert Interviews. We interviewed the Product Marketing Manager of my previous company, along with the Head of Product and quickly discovered that there is a clear gap in the market for small online businesses looking to find the right influencers in the niche to help with influencer marketing.
Ideation
We came up with multiple solutions that targeted the problem from different places including a platform for influencers to monitor their finances and a way for small businesses to bid for ads within YouTubers videos (this post is brought to you by Raid: Shadow Legends). However, given our expertise in machine learning, we decided to work on a matchmaking platform that uses NLP to go through an influencers community and identify the niche that they work in as well as the authenticity of the connection they have with their audience. This is where the process of the design sprint is the least helpful. The process is mainly geared to testing a solution with testers, rather than coming up with net-new business ideas, so we were kinda on our own at this stage.
Brainstorming
The process, in our opinion, really shines in giving the team a structure to think of ideas independently and then brainstorm collectively. We would all have time by ourselves to sketch out the different steps in the business model and then we would collectively decide what our approach would be. We particularly liked the ‘Crazy 8s’ exercise that forced us to come up with 8 different layouts to our product in 4 minutes and the ‘Lightning Demos’ where we all showed our different features from some other products to help inspire our own product.
Mockups
This whole process is designed to test a solution as soon as possible with users, so we knew that we would not write any code yet. Luckily, the market is teeming with no code UX mock tools. We ended up using Framer.com and we liked it a lot. It allowed us to put together a reasonably high fidelity mock of the product in a day! The key here is to focus on testing the part of the customer journey where the assumptions you made about a customer’s problems come to fruition. For us, our key assumption was that customers would pay to quickly and easily find the right influencers to reach out to for sponsorship without spending hours trawling through hashtags on social media. So our mocks were focused on the process of influencer discovery.
User testing
We approached small businesses in our network to ask for an hour of their time to walk them through our prototype. The book recommends 5 user tests because you experience diminishing marginal utility quite quickly after that. We wanted to keep our user testing base as diverse as possible, even within our sample size of 5 user interviews. We didn’t just want a bunch of tech bros like us agreeing with everything we were saying and being more interested in which NLP algo we were using. So we interviewed men and women of different ages, races and industries to get an understanding of the problem space and if our prototype helped them. We also posted on Reddit and got 2 user testers from there!
The most important thing to be aware during the user testing is to not ask leading questions. You want the users to tell you about their problems, and what they think your prototype is trying to do rather than asking questions like ‘Is this page too busy?’ or ‘Does this page help you find the right influencers?’. Also, never ask users for solutions, only problems.
Results and Next Steps
We got positive feedback from our user tests! Almost universally, they all agreed that influencer marketing was something that they have tried or have been wanting to try, but the amount of time it takes means that the ROI is quite challenging to justify. Indeed, 2 of the 5 people we interviewed said that they would be very interested in signing up for our beta! However, talk is cheap. We knew that we would only be able to actually know that this idea was worth pursuing if they paid for it. So we asked them for payment and within 30 minutes our Venmo was dinging. We got 2 paying customers with our prototype and knew we had something to work towards. We now have a landing page, Matchfluence and are driving traffic to it to measure the need for a product like this from a quantitative perspective. We do not have much design experience, so any feedback on the landing page and the problem space is welcome! Also don’t be shy to sign up for our beta if you want to explore influencer marketing
I hope you enjoyed this post and I hope this comes in handy if you are considering a design sprint for your business!
Background
Two friends and I (tech backgrounds) decided to spend a week trying to test out a business hypothesis. I had done a design sprint at a previous startup when we were trying to pivot and thought this would be a great structure for us to come up with something concrete by the end of a week. We decided on a general problem space of helping small e-commerce stores based on the thesis that as people lost jobs and had to work from home post covid, the amount of people starting one man (person) online shops would increase. Shopify’s latest numbers are proof. My last startup worked with Twitch influencers as the main customer acquisition channel, and I knew first hand what a disaster the process of working with influencers/content creators was/is.
Problem Space
We used a modified version of the process, found here https://www.invisionapp.com/inside-design/design-sprint-2/ and started with the Expert Interviews. We interviewed the Product Marketing Manager of my previous company, along with the Head of Product and quickly discovered that there is a clear gap in the market for small online businesses looking to find the right influencers in the niche to help with influencer marketing.
Ideation
We came up with multiple solutions that targeted the problem from different places including a platform for influencers to monitor their finances and a way for small businesses to bid for ads within YouTubers videos (this post is brought to you by Raid: Shadow Legends). However, given our expertise in machine learning, we decided to work on a matchmaking platform that uses NLP to go through an influencers community and identify the niche that they work in as well as the authenticity of the connection they have with their audience. This is where the process of the design sprint is the least helpful. The process is mainly geared to testing a solution with testers, rather than coming up with net-new business ideas, so we were kinda on our own at this stage.
Brainstorming
The process, in our opinion, really shines in giving the team a structure to think of ideas independently and then brainstorm collectively. We would all have time by ourselves to sketch out the different steps in the business model and then we would collectively decide what our approach would be. We particularly liked the ‘Crazy 8s’ exercise that forced us to come up with 8 different layouts to our product in 4 minutes and the ‘Lightning Demos’ where we all showed our different features from some other products to help inspire our own product.
Mockups
This whole process is designed to test a solution as soon as possible with users, so we knew that we would not write any code yet. Luckily, the market is teeming with no code UX mock tools. We ended up using Framer.com and we liked it a lot. It allowed us to put together a reasonably high fidelity mock of the product in a day! The key here is to focus on testing the part of the customer journey where the assumptions you made about a customer’s problems come to fruition. For us, our key assumption was that customers would pay to quickly and easily find the right influencers to reach out to for sponsorship without spending hours trawling through hashtags on social media. So our mocks were focused on the process of influencer discovery.
User testing
We approached small businesses in our network to ask for an hour of their time to walk them through our prototype. The book recommends 5 user tests because you experience diminishing marginal utility quite quickly after that. We wanted to keep our user testing base as diverse as possible, even within our sample size of 5 user interviews. We didn’t just want a bunch of tech bros like us agreeing with everything we were saying and being more interested in which NLP algo we were using. So we interviewed men and women of different ages, races and industries to get an understanding of the problem space and if our prototype helped them. We also posted on Reddit and got 2 user testers from there!
The most important thing to be aware during the user testing is to not ask leading questions. You want the users to tell you about their problems, and what they think your prototype is trying to do rather than asking questions like ‘Is this page too busy?’ or ‘Does this page help you find the right influencers?’. Also, never ask users for solutions, only problems.
Results and Next Steps
We got positive feedback from our user tests! Almost universally, they all agreed that influencer marketing was something that they have tried or have been wanting to try, but the amount of time it takes means that the ROI is quite challenging to justify. Indeed, 2 of the 5 people we interviewed said that they would be very interested in signing up for our beta! However, talk is cheap. We knew that we would only be able to actually know that this idea was worth pursuing if they paid for it. So we asked them for payment and within 30 minutes our Venmo was dinging. We got 2 paying customers with our prototype and knew we had something to work towards. We now have a landing page, Matchfluence and are driving traffic to it to measure the need for a product like this from a quantitative perspective. We do not have much design experience, so any feedback on the landing page and the problem space is welcome! Also don’t be shy to sign up for our beta if you want to explore influencer marketing
I hope you enjoyed this post and I hope this comes in handy if you are considering a design sprint for your business!