NOVEMBER 7, 2017
5 min

10 Product Management takeaways from Intercom


I checked out the excellent Intercom on Product Management and here are 10 insights from it that can help you build better products and become a better Product Manager:

1 - Audit your feature usage

If you don’t know how many people are using your features, and how often, then you don’t know where to improve your product.

Use an analytics tool (Mixpanel, Intercom, Google Analytics etc), SQL or whatever works for you, to plot all your features (excluding administrative features as password resets) on two axes: how many people use a feature, and how often.

Here's a typical feature audit example from Intercom:

intercom-1The core value of your product is in the top right area because that’s what people are actually using your product for.

If most of your users are using one feature a lot and the others are not being used, then you have a bloated product and a competitor focusing on that one feature could quickly overtake you.

Here's an example of ideal feature usage:


2 - Be careful with customer priorities

If you ask customers if they would like you to add feature X, they will probably say yes! This is a no-loss question for them. Within a company, Product Managers are used to trade-offs because resources are finite. However, customers do not care about internal struggles and nor should they.

To get a better idea of what customers really want try making them choose between features e.g. Would you rather we added more search options or added this integration for you?

I’m not a big fan of asking customers to prioritise a list of features because; a) that’s work for them to do to help you, and b) they will probably all come back as ‘Must haves’ which gives you no insight.

3 - Say No to features more

There are no small changes in Software Development. Seemingly innocuous or superficial features like character limits can impact existing features or disrupt loyal users. Even copy adjustments can change the perception of your product. Also, everything adds technical debt and complexity.

"Scope grows in minutes, not months."

It’s easy to keep developers busy with a backlog of feature additions that are easy and ‘no-brainers’ to add. But your product will get bloated one careless decision at a time. If all you do is add new features then your product will be miles wide and inches deep — the dreaded Jack-of-all-trades.

4 - People only use your product to solve a problem

Personas are great as they help you understand your users and markets. However, products are not prescribed to people based upon their demographics, frustrations, good days, emotions, [insert relevant persona metric here]. Take it back to basics — people use your product to solve a problem that they have. Even when purchases are seemingly based completely on brand loyalty, the user is still solving a problem e.g. I want to wear Nike to look cooler in front of my friends.

The Jobs To Be Done Framework is good reading for Product Managers who want to understand why people use a product (and that should be all of you…)

5 - The 3 ways to make products people want

Plenty of software gets created that is not useful to anyone. If you think you’ve identified an opportunity for a product or feature that can solve a customer pain-point then it should do one of these three things:

  1. Remove steps/make it easier: Deliveroo reduced the number of steps needed to get a meal.
  2. Make it possible for more people: Thanks to Snapchat, Instagram, Youtube etc vlogging is now available to everyone, not just people with production companies.
  3. Make it possible in more situations: Mobile Banking apps let people manage their finances on the move without needing to visit a bank during working hours.

“It’s easier to build things people want than it is to make people want things”

6 - There are 3 ways to improve a feature

  • Deliberate Improvements: Your aim is to make a feature better

These improvements make a feature better for the people who are already using it e.g. easier to use, faster. Deliberate improvements on core value features (see lesson #1) are high risk/high reward because you can delight your entire user base or break your core workflow.

  • Frequency Improvements: Your aim is to make a feature be used more often

Adding more options or more hooks into a feature e.g. notifications might get existing users to use the feature more. The Hooked Model by Nir Eyal is an excellent guide to help with Frequency Improvements:

  • Adoption Improvements: Your aim is to make more people use a feature

These improvements target people who aren’t using a feature. Examples include better onboarding, in-product help and viral gamification. The 5 Why’s technique is a great way to get to the root cause of why someone might not be using one of your features.

7 - Have an acid test to approve new features

We’ve discussed that you should not continue to add features to your product without seriously considering the value they add. As a Product Manager, it can be hard to keep multiple stakeholders happy with a finite development team at your disposal. Having a list of criteria that any feature request must pass before it is eligible for prioritisation can help make your decision making transparent and unequivocal.

Intercom list 10 questions that a feature must score straight yes’s on. You can check their list in the book but here are my favourites and some of my own:

  1. Does it fit in with the product vision? (Gospel! Always move towards your vision)
  2. Does it grow the business? (A yes here can conflict with #1 so be careful as growth should always be front and centre)
  3. Does it contribute to a predefined, measurable goal? (e.g. adoption. This will stop you from accepting ad-hoc requests that don’t support #1 or #2)
  4. Will the benefit out-way the support and maintenance costs? (think ahead, technical debt piles up)
  5. Do we have the competencies to build it well? (releasing early is not an excuse for a botched job)

8 - Prioritise using ROI

Takeaway #4, there are no small features! (see part 1). With that in mind, focus your roadmap around features that offer value for money, because everything costs money. Plot your backlog items on a quadrant with User Value and Effort as the axis.

Intercom suggests plotting the ROI of features like this:

intercom-4Users only care about items on the right-hand side of the quadrant so make sure your roadmap doesn’t have a long period with none of these items. Of course, it’s necessary to do work from the left-hand side e.g. Architecture improvements, but spread these out so users can see constant improvement in your product.

Items in the bottom right are the Low Hanging Fruit, this is usually the rarest category as the benefit disproportionally outweighs the investment. Use these to balance out periods where you need to add changes from the left-hand side.

Here is another way of prioritising features based upon ROI (taken from Product Management is like roulette, except the house doesn’t always win):

intercom-59 - Roll out strategy matters

This one depends on the stage of your product/company, the size of your user base and the nature of your workflow. Early-stage companies with few users and non-essential workflows can consider immediately launching features to all users once their test is complete. However, if you are an established company with lots of users who rely on your product for mission-critical tasks (think B2B) then a phased rollout is safer.

Intercom using the following approach: 
Team testing > Company testing > Beta testing > Full rollout

10 - Communicate about new features properly

Designing, building and testing features is engaging, so it’s easy to get addicted to doing this and never look up from the trenches. Once the code is shipped it’s easy to forget about it and move on.

"A feature without engagement is not a feature, it's technical debt."

Companies and Product Managers do not win by having an ‘all encompassing’ product with lots of features. Engagement - winning!

Here are some ways to properly communicate a new feature so users become engaged:

  1. Show how it benefits them — Gospel! Nobody cares how smart your product is, they care about how it can help them.
  2. Show how they can use it — Don’t patronise…but never assume that they know it as well as you, because you built it.
  3. Announce it in the right place — Communicating inside your product is better than outside via email. If you want a user to try your new filter feature then they are far more likely to do that if they are already filtering information when you alert them.
  4. Announce it at the right time — If a user has just joined your product, the first onboarding popup should not be about data export as the user hasn’t got any data yet. Map out your user journey and introduce features gradually at the relevant point as your users advance.

Here's an example of good communication along a user journey:


There are plenty of great lessons for Product Managers in Intercom on Product Management and here are my top 10:

  1. Audit your feature usage
  2. Be careful with customer priorities
  3. Say No to features more (there are no small changes!)
  4. People only use your product to solve a problem
  5. The 3 ways to make products people want
  6. There are 3 ways to improve a feature
  7. Have an acid test to approve new features
  8. Prioritise using Return on Investment
  9. Roll out strategy matters
  10. Communicate about new features properly
Helping you from idea to launch to scale.

Keen to find out more? Get in touch!

We'd love to chat to you.