Product Management | 5 min read

Ways to Improve a Product


Kaizen is the philosophy of continuous improvement. Web businesses searching to find product market fit all follow some variation of Kaizen whether they know it or not.

Shipping code doesn’t mean that you’re improving anything. Similarly, you can make undeniable improvements to parts of your product and get no response or appreciation for it. It all comes to down to type of improvements you’re making.

The two most popular ways to improve a product are to add new features, or to improve existing ones.

Adding New Features

New features expand the scope of the product, often making a big marketing splash, getting a version bump, and resulting in some press releases. Often the fanfare attracts new customers and new use-cases for the product. Typically, new features are the only improvements that outsiders (i.e. non-customers) will ever hear about.

New features are risky. You have to be very confident they will be valued, as they’re like children; you have to love them and support them no matter what.

Ask your customers “Would you like a [Calendar|TimeTracker|Gantt Chart]?” and they’ll reply yes. It’s a one-way “something or nothing” offer. They haven’t had to make a trade-off between competing priorities. This leads to customers saying they want stuff that they don’t really want.

Asking your customers “Would you rather that we made the product much faster, or that we added more labelling features?” and you’ll get a different answer. Everyone values speed. (Sidenote: Don’t you wish the Gmail team did this)

Improving existing features

You can improve an existing feature in 3 different ways: you can make it better (deliberate improvement), you can change it so customers use it more often (frequency improvement), or you can change it so more people can use it (adoption improvement).

A) Deliberate Improvements

Covered previously, this is when you know why customers use an existing feature and what they appreciate about it. A deliberate improvement seeks only to make it better in ways that will be appreciated by the current users.

Use when: There is a feature that all your customers use and like, and you see opportunity to add significant value to it.

B) Frequency Improvement

These are improvements that hope to get a customer to use the feature more often. Adding more items to an activity feed, or more options to a search tool means that people read it more often, or use it for more tasks each day. This type of improvement can turn a once-a-week feature into a every day feature.

Use when: There is a feature that the majority of your customers use infrequently, and you believe that using it more would be a benefit to them.

C) Adoption Improvements

Adoption improvements target those who don’t use a feature. For example, if you have a calendar feature that only half of your user base is using, an adoption improvement aims to win over the remainder. By adding ICS file import, or a Google Calendar connection, you don’t improve it for all users… but you make it more likely that the non-consumers will now adopt it. When you hear phrases like “I’ll use this as soon I can…”, then an adoption improvement might help.

Use when: There is an important feature that a good chunk of your users have yet to adopt, and you see some obvious integrations or changes that will make it easier for them to get on board.

I write about the different ways you can improve your application to make a point. Not all improvements will be appreciated by all users, and it’s important to understand where the opportunities for real improvements are within your product.

Improving Intercom

shot of comment

Unless you specifically make a point of calling them out, most improvements go unnoticed by outsiders, and also by users. This leads to comments like the one above.

So let’s talk about what we’ve improved in Intercom, here’s 5 recent changes we’ve made.

1. New Message Types

shot of comment

Intercom’s previous message style was a pop-up to start a conversation. There are times when that doesn’t cut it, for example:

  1. You want to send a message, not start a conversation.
  2. You don’t want to interrupt your user.
  3. You want to send a rich message with images, etc.

The effect of this was that not all of our customers were using in-app messaging, and those that were, weren’t using it for everything. We added notifications (Growl-style pop-ups shown above) and announcements (large pop-ups without reply) to address this.

2. New Results View

shot of Results view for message analytics

With all these new messages being sent, we wanted to give our customers an easy way to see how effective each message is at driving replies or click throughs. This led to a re-design of our message results view, to put this information visible anywhere a message is.

3. New Filter Bar and Segment bar

shot of filter & segment bar

Filtering is how our customers discover who is using their product and how. Our previous filter didn’t scale for complex queries, or for large amounts of data. In addition, there were far too many clicks involved in simple, frequent queries. To solve this we re-designed and re-positioned filters as shown above. Most common queries are much quicker, and more complex ones are far easier to configure.

4. User notes

shot of user notes in Intercom

Large teams with large user bases need some way to share customer insights that come out of calls, customer development meetings, meet-ups, and other offline activities. Tagging users is great for sharing one discrete piece of knowledge (e.g. Wants Android App), but for telling a story with context, notes are far better.

5. New Changes Site

shot of changes site

We’ve also rolled out lots of speed improvements (via AJAX and PJAX). We’ve made our Intercom JS behave well for single page apps and mobile web applications. We’ve added an assignment bar when viewing conversations, and made lots of bug fixes. We can’t keep listing all these changes, and they don’t always merit announcements, so we built a changes site that records them. It’s live and open to the public too.

Continuous Improvement

In their early years new web businesses have advantages over the incumbents. They can move quicker and adapt faster, without as much technical debt, legacy features, compatibility issues, or big value customers restricting their movement. Often this speed and agility can cause start-ups to pivot like headless chickens, rather than focusing on improving their product in meaningful ways. The challenges of the product manager are two-fold; firstly, finding improvements that will benefit the business and its customers, and secondly, ensuring that these improvements don’t get lost on a whiteboard somewhere, and actually make it out the door. Because if there’s one thing that’s true for start-ups web products, it’s this:

quote by @benmilne