As a startup founder, what do you use - to find product-market fit?
Venkat
22 replies
How do you establish if you are building a product that fulfils a need?
How do you know you are selling it to the right people?
Is it driven just through conversations with customers, or are there other inputs that helps you decide what you are going to do next
Replies
Stephane Ibos@stephane_ibos
Conversation is always a very good way to identify if your product is really useful or a 'nice to have'. It also helps you identify frictions (at every level: on-boarding, usage etc.).
When an actual conversation is not possible, you have to be a KPI-freak. Measure everything: click rates, bounces, interactions, pauses. You should also measure behaviour (I recommend Hotjar for this - it records everything your user does on your platform). Important KPIs of course: usage, renewed usage, churn.
If you have plenty of folks who try and leave to never come back, you know you do not have a product-market fit. Be it because of your product, or because you do not target the right audience.
Run trials to see if changing audience improves things. If not, question the product itself.
You can organise focus-groups. Quite useful as you can really narrow your analysis down to any particular aspect you want.
Share
@venkatasubramanian_k Yes you do need to identify your key personas to gather the right ones in a focus group. Unless you want a broader 'discovery' group (KYC activity = Know Your Customer) but that has a different set of objectives.
We used to pay for focus groups. Agencies exist that can help you set one up. It's not overly expensive (total cost including 'time paid' for attendants was around $1000 if I remember well). In my experience totally worth the expense.
@stephane_ibos thanks for your insightful response, running analytics on product usage will sure help us uncover potential points of failure and opportunities. How do you suggest founding teams can create these focus groups? discord or slack groups or direct messages etc.? Sounds like we should also be thinking about who should/should not be a part of the focus group.
I think communicating with customers (current or potential) is definitely the way to go here. A lot of product ideas come from people sharing their pain points and experiences where something was missing. Looking into competition always helps too, we can get a lot of information from reviews and opinions on what works and doesn't with a similar product to see if ours will remedy the problems and offer a better solution.
@dslabglobal_ana Definitely. Customer conversations are a must. :)
Discussing with potential customer helps a lot as previously mentioned.
Tracking the activation rate and retention are the best way to check product market fit.
How many of the website visitor signup to try your product?
How many of the signup completed the on-boarding and experienced the first value point?
How many user are still there 10 days later?
@fabian_maume thanks. There are tons of products out there that help us do this. In your experience, which one do you recommend? Like GA, Hotjar, Heap etc?
@venkatasubramanian_k I would recommend:
- GTM to track all your events and triggers other analytics tags
- GA is always nice to have. Main issue is it doesn't allow to analyse data a posteriori which can be an issue for startups
- Some event based analytics like Mixpanel or Amplitude to be able to dig in data a postreriori.
- Some qualitative tool for screen recording & heatmap. I would recommend Smartlook for that.
You can schedule a webinar/demo of your product, share the invitation on social media (and even provide some discounts to the participants) and ask for a feedback on the product there. Later you can share the recording of the video to get even more feedback in the comments under the video
@slava_bobrov that's an intersting idea. Will definitely try this. Do you use any specific product for this?
@slava_bobrov We tried this method, and landed up with our first paid account with 10 users:) but for feedback we called up our target users individually to get a better understanding of their needs
@nilova - looks like a method that works. Congrats. :)
1. What kind of questions do you specifically ask when you ask for feedback?
2. Let's say in a month's time you have 100 users. Talking to them individually is going to get difficult. How do you envision keeping up with the aspect of getting qualitative feedback about your product. Will you be looking at speaking to a select sample of customers? Or have an automated way to reach all customers? Or both?
@venkatasubramanian_k I use Eventbrite, also posting a zoom meeting registration works as well
Ideally you have the problem or really understand it. No one invests in an pizza place founded by a guy who doesnt know what pizza is or even likes it. So be close to the problem. If close to the problem, then you can likely find more of you. But finding more of you meaning people that you think would care about it isnt good enough- you need to expose the problem to them and reveal your solution without biasing them along the way. So assume you do that too- then its all speed, speed to a better narrative, speed to a first prototype, speed to the first conversation with a real customer in front of your prototype, speed to the realization you should of done x vs y etc etc.
@floozyspeak I couldn't agree more on the speed part. The problem I am trying to solve is optimising the time to find pmf. Initiated this discussion because I want to understand how founders usually go about it. I would be really excited if you want to try it out and provide feedback if there is any. Thanks :)
@venkatasubramanian_k we organised with the admin of the groups to post TypeForm surveys to their page and invite responses. The responses feed directly into the AirTable base or alternately you can just build the survey form in AirTable itself.
A few hours to build the initial template which you can see here (https://airtable.com/invite/l?in...) then probably a day or so to get all the segmentation and analysis right once we had the data.