Development continues with Retainable, but as I was anticipating, I only wish that it could move along faster, with fewer technical challenges. But that’s not a part of this start up thing, as I knew going in….
But we have hit a few small unforeeseen speed bumps that changed the direction of product development, and I think ultimately for the better…..
In short, my initial plan was to use Mixpanel and analyze historical user data to identify upcoming churn events. This seemed to make sense, but i still had lingering doubts…..could users just do this with mixpanel now anyway? why would they really pay for another app to just interpret what they are already paying for? how many of my actual customers were actually using Mixpanel anyway?
My thought was that Mixpanel would just be the first data set to use to help interpret the in app behavior of customers. but it is very one dimensional, only based on their actual data.
But there are many other data inputs that would be useful in creating a holistic view of the customer: what about the qualitative feedback from their surveys or NPS scores? Their engagement with email campaigns? the success, or lack thereof, of their onbaording campaigns? NPS scores? upgrades or downgrades? there are so many data sets that are informative about a customer, and when aggregated can give a very telling picture of how engaged they are with the product. and from there, how likely they are to stay a customer.
Regardless, the few Mixpanel challenges with the mixpanel integration put a halt to my singular focus on a Mixpanel integration: users need to have a paid plan in order to export user data. and even in doing that, they can only export for selected dates. or if it was going to be a continuous stream of data via data pipeline, then it would create incremental costs to the Retainable customer’s mixpanel account. That would be a terrible experience. it would be like watching Netflix marathons in the 1990s, where there would be an incrememental cost for every minute of dial up bandwidth used.
So it was a good juncture to pump the brakes on the initial plan of the Mixpanel integration and think of a solution that would allow for more customer feedback, and interaction, from the start. After all, it would be pretty annoying if i had a handful of beta testers open to using Retainable, only to decline the offer because they were using a different tech stack.
Instead, we are making a slight shift to be able to ingest all data in a CSV format. This seems intuitive now, allowing us to use data from their email platform, CRM, product analytics, social media, etc to make an informed decision about their future churn.
Taking A Step Backwards. Or A Step “Down The Stairs”
I recently came across the notion of a “stairstep approach” of Rob Walling and popular in the bootstrapper space. In short, the premise is to progress in incremental stages by starting small—selling one-time products like extensions or plug ins, and after validating an idea, honing skills, and improving product-market fit, then proceed towards a recurring revenue subscription.
Why jump in to building a robust product if there is no certainty that there is a need and potential customer base? Time and money can be spent in more pragmatic ways.
This stairstep approach has been weighing on my mind as I uncover all of the potential challenges with product development, sales, and marketing. Are there things that I am doing now that I could scale back, or not do at all? I thought that I was thoughtful in my prioritization, but maybe I wasn’t ruthlessly prioritizing.
An example of this was adding Stripe to our check out flow. I thought that it should have been a plug-and-play solution. After all, I thought that it was popular particularly for it’s developer-friendly nature. But there were a few challenges that took a few days to fix. But do we even need Stripe at this point? Could I just use a PayPal form, or take a credit card number by hand? though not an ideal solution, yeah, I could probably do that in the very beginning….which just means that I still had some fat to trim on the initial version of the product I am building.
Did I Bite Off Too Much Too Soon?
The whispers uncertainty and self-doubt about the idea, and my proposed solution, started to creep in as I wasn’t making the progress that I was hoping to. I mean, an MVP should be completed in two weeks, right? I mean that facetiously, but it’s all relative…there are some solutions that could have been tied together that would validate a proof of concept and save some time.
This could be achieved with pre-sales: asking potential customers if they would pay for a proposed solution, even though it is not developed yet. Generate enough confirmed customers, and there is a pre-defined set of customers to launch to.
One downside with pre-sales is the unsatisfying void between getting a pre-sale and delivering an actual solution.
So I want to split the difference, and shift towards a “productized service” or agency model in the short term.
A productized service is basically taking an agency model of doing a service for a client, but with one key differentiator: there are a set of specific deliverables that are agreed upon every month. The benefit is that it sets clear expectations, and avoids lots of back and forth, one off client calls, scope creep and the time suck that can come with an agency or consulting model.
While I am building out the MVP of Retainable, doing a productized service would provide several important benefits:
- Customer Validation – I know that customer churn is a problem for software companies. But if I’m really honest with myself, I don’t know exactly how big of a problem it is. This is maybe the classic case of “scratch your own itch”, where I experienced the challenges of churn for SaaS businesses, but then I hear about companies that churn out 1-2% of their user base a month. Churn is invevitable, so do these companies take their lumps and move on to bigger priorities? Or do they pinpoint it as a critical problem for their company to solve? I think that I know the answer, but I am really just making some broad swath assumptions.
- Generating Revenue – I am bootstrapping Retainable, and that is my goal for the foreseeable future. But if I am going to build a pure play SaaS, then it can be a long time before a product of value is built. I can not build it myself, so in the interim I am working with a contract developer (ie a very real monthly expense) until I can find the right fit with a technical co-founder. Which brings me to the next point….
- Finding Product Market Fit – How much of a need is Retainable solving? How much process, and insight, do companies have on their monthly churn numbers? My guess is that there is a bit of a band aid approach to do basic blocking and tackling: monitor monthly churn, create some onboarding emails and customer personas, maybe an NPS survey or two. But there is a limited process and understanding to churn. and it’s not for any particular reason except for lack of bandwidth. I think that Retainable can bridge that gap and provide the insights, but there is not a dollar to back that claim….yet. Moreover, what is the profile of those customers? B2B SaaS, agencies, marketplaces, B2C apps? I’m still not positive on this.
- Refine the sales process – I have not made an actual sales call for Retainable yet. I have done a handful of customer discovery and ideal validation calls, but I never pushed the proverbial pen across the table and asked for the sale. So I am not sure what the objections may be, what the questions may be, what the alternative solutions may be. These will all come to light if I am selling a productized service.
- Save On Development Costs – In some ways, building Retainable feels like embarking on a maze with no map, flashlight, or guide. It’s poking around until there is some semblance of a path forward. That’s fine, all part of the process. But it is not fine, and sustainable, if I am working with a developer who is being paid by the hour, and ultimately building features that never see the light of day. If I can avoid development hours that a fruitless, I damn sure going to do everything I can to find that path! A productized service is a good interim solution to build the absolute minimum and guide future development based on customer feedback.
- Build A Business – And by that, I convince some strangers that I can help them solve a problem that they have. And that I should be paid for that. This, building a business, is basically the summation of all the points above. Whatever it is called—software as a service, productized service, agency—is all irrelevant if there is no customer paying money.
Where does this leave me?
We are going forward with product development, but with a focus on creating accurate and scalable machine learning algorithms. This of course is another challenge, as every business has different metrics of importance, so it isn’t going to be an easy task to automate the pre-processing and analysis of various data sets for various companies.
Personal Update
It’s been hectic few weeks: two birthdays (Aya and mine, just two days apart), Mother’s Day, and preparing our place to rent out on Airbnb.
We are going to spend the summer in Spain and Portugal–places that we have wanted to visit for a while but have not had the chance. It is an exciting adventure for the summer, and one of the primary goals: having the ability/opportunity to be location-independent as a family while continuing to work.
Here’s a picture from the highlight of the month, at Aya’s first big birthday party (turning 3!) with Miss Sparkles (she’s in the gold top):
Recent Comments