When applied to your product/service, I’ve found an approach that will give you the clarity you need to take action that leads to growth.
But before we dive into the approach, here is some background.
I’ve spent half a decade working with startups on growth and have observed many startup teams get lured to destruction by the appeal of vanity metrics.
Investors and founders lean on metrics to determine how the business is doing. We also use metrics to lie to ourselves. E.g., Your latest product post had 120K more impressions than the previous post but contributed to 0 sales conversions. Aka, busy making progress that doesn’t matter to your core objective.
A signal for activity rather than results.
So what’s the approach?
The approach focuses on understanding state changes among your product/service users. To serve this up, we will look at a concept known as Finite State Machines.
A few caveats before we continue. This approach assumes;
- You already have users/customers engaging with your product/service.
- You have already established some “Fit” (you are making things that people want).
- You are targeting a specific kind of user to grow your user funnel (making people want things).
What if I’m starting from scratch?
Try to get at least 100 to 1000 users (B2C) or 5 to 20 users (B2B) before you apply this approach. There is nothing special about the B2C or B2B numbers, they are just a range that can give you some statistical significance.
What is a Finite State Machine (FSM)?
FSM is a mathematical model of computation used to design logic circuits and computer programs. Professor Elaine Rich and All-Turtles founder Phil Libin have videos that inspire my use of this concept.
FSM concept consists of the following:
- States (circles)
- Transitions between the states (arrows)
- Inputs and outputs
Example: Vending Machine
- States = Empty, Waiting for payment, Dispensing product
- Inputs = Insert coin, Push button
- Output = product Dispensed
Why do Finite-State Machines matter?
Finite-State Machines make it easier for us to understand and predict behavior.
An oversimplified example of an air conditioner FSM:
An oversimplified example of Traffic Lights FSM:
FSM provides a simplistic way of looking at your product or service and identifying the critical transitions and states that will grow or kill your business.
How to apply Finite State Machines to product growth:
- We have approx eight billion people available in the world.
- Some of these eight billion people will be your ideal target customers or fall into your target segments.
- We can use this information to establish two core groups of people concerning your product/service:
Groups:
- People that have used your product/service (UPS)
- People that have never used your product or service (NUPS)
The UPS group has four types:
- First-time users
- High-value users
- Low-value users
- Inactive users
Remember, this approach will not be of much help if you have no users. Get your first 100 (B2C) or 5 (B2B) through the First-time user state before you can identify the appropriate user classifications.
What is value?
Value is defined from the user’s perspective. What usage scenarios indicate importance to the user?
The knowledge of your ideal customer and their pain points should clarify this for you. e.g.,
- Restaurant visitors who visit at least two times a week.
- The number of hours spent playing a video game.
- The number of products they purchase from a single site visit.
- The number of repeat orders they make in a month.
Or whatever use cases signal high usage and relevance to the user within a reasonable, measurable period.
To define who is a high-value or low-value user, you will need to:
- Make predictions about their behaviors
- Estimate their usage scenarios, activities, and associated frequency
- E.g., the number of sessions a month
– High-value (30+ sessions)
– Low-value (29 sessions or less)
Example Facebook:
- Creating a Facebook profile and browsing/stalking users might be a low-value usage scenario.
- Creating a Facebook profile and page listing products and linking your WhatsApp to handle leads and queries may be considered a high-value usage scenario.
Ultimately, your ideal customer profile and their core objective should determine or at least highlight the high and low-value usage scenarios.
Example: Netflix
If value, for the user, is how much entertainment they get in exchange for a $ 15-a-month subscription, then a high-value user could be classified as follows.
Low-Value user (example)
- Launches app daily
- Browses catalog
- Adds shows to the watchlist
- Watches 1 show weekly
- Watches 40 minutes a week
High-Value user (example)
- Launches app once a week
- Watches 3 shows and 1 movie weekly
- Watches 10hrs every weekend
You could classify users based on this to track the volume or % of users in each Finite-State and observe what triggers a transition to a new State. E.g., the launch of a new series might move the needle for some users.
After defining the criteria, you can look for aligned incentives in the business model. i.e., In the above example, high-value users (40hrs of streaming) and low-value users (2hrs 40mins of streaming) deliver the same revenue to the business, $15 per month, signaling the potential to optimize the value extraction in your business model.
Goal: Get as many high-value users as possible.
Objective: Increase the number of high-value users by 15% every 30 days (or whatever period makes sense for the business).
14 User States & how to use them:
There are 14 possible user states.
- Never used product/service uses product/service for the first time
- Never used product/service never uses product/service
- The first-time user stops using the product/service and becomes an Inactive user
- The first-time user becomes a low-value user of the product/service
- The first-time user becomes a high-value user of the product/service
- Inactive user stays Inactive
- An inactive user starts to use the product/service and becomes a low-value user
- An inactive user starts to use the product/service and becomes a high-value user
- The low-value user stops using the product/service and becomes an Inactive user
- The low-value user increases their usage and becomes a high-value user
- The low-value user maintains their low usage and stays a low-value user
- High-value user decreases their use and becomes a low-value user
- The high-value user stops their usage and becomes an Inactive user
- The high-value user maintains or increases usage and stays a high-value user.
You need to keep an eye on two kinds of State changes.
- Uphill State changes
- Downhill State changes
Uphill State changes (step up)
Have a pulse on all the uphill state change triggers and be able to replicate the triggers, especially with newly targeted First-time users.
- First-time user to high-value user
- Low-value user to high-value user
- Inactive user to low-value user
- Inactive user to high-value user
Downhill State changes (step down)
Have a pulse on all the downhill state change triggers and be able to remediate them before they reoccur, especially with newly targeted First-time users.
- First-time user to inactive user
- Low-value user to inactive user
- High-value user to low-value user
- High-value user to inactive user
Takeaways:
Plug Your Leaking bucket: (fix this!)
When (NUPS) Never-used your product/service becomes a First-time user and then becomes inactive-user.
- You need to set user expectations correctly. Overpromising?
- You need better or clearer product/service messaging.
- Product/service has a bad user experience
- There is too much friction or effort required to enjoy the value proposition
Fix your sales and marketing only after establishing a statistically significant bounce rate. If only two people have used your product or service, concentrate on getting to a hundred users before you start incorporating their feedback or making iterations.
Low-value churn: (potential distraction)
The low-value user stops using the product/service and becomes inactive.
You might learn some things from these users, but focusing your efforts on high-value users would be better. An exception might be if the low-value user activity is a crucial input into a network effect or a flywheel that influences high-value user behavior.
High-value churn: (fix this fast!)
A high-value user that stops using the product/service and becomes inactive.
These are the vital signs of your business. You will uncover the early warning signals and critical friction points: quick wins, turnaround scenarios, and win-back opportunities.
Remember to separate high-value churn from low-value churn and bounce rate in your efforts and prioritize high-value churn, then bounce rate, and only once those are handled can you address low-value churn.
Raising the Dead is no easy business and should be an act of last resort.
- Trying to turn inactive users into low-value users.
- Trying to turn inactive users into high-value users.
- Trying to turn low-value users into high-value users.
Ideal Customer Profile Fit
When a First-time user immediately becomes a high-value user and stays a high-value user.
You need to understand these users and identify how to reach more of them as cheaply as possible.
Upgrade me, please
The low-value user increases their usage and becomes a high-value user.
Cultivate: (retention)
High-value users that maintain their high-value usage
For all the states:
- Understand their respective user journeys.
- Determine the operational cost required to get users from one state to another.
- Avoid or fix the most costly journeys.
- Master and replicate the most profitable journeys.
I hope this was helpful.
Happy Growing!
Sources & Inspiration:
- For those that want to geek out on the structure of state machines https://ptolemy.berkeley.edu/projects/chess/eecs124/reading/LeeAndVaraiya3_4.pdf
- Tara Hunt – Metrics are a lie https://www.youtube.com/watch?v=5qM3G-2HDT4
- Elaine Rich – Finite State Machines – https://www.youtube.com/watch?v=lh2onWfBrxk
- Phil Libin’s founder institute video https://www.youtube.com/watch?v=3XumZ_LT_ZE