Northwestern Mutual becomes data-driven with Heap

About Northwestern Mutual:

Northwestern Mutual provides consultation on wealth and asset income protection, financial planning, and private client services. Its products include life insurance, disability income, long-term care insurance and investment advisory services.

We asked Joe Pfeiffer, VP of Product, what he likes best about Heap.

Joe Pfeiffer

Joe Pfeiffer, VP of Product

INTERVIEW:

What products are you responsible for?

I’m the VP of Product for our consumer products group. I run a team of about 15 product managers who build the brand marketing website at NorthwesternMutual.com, and our client website, which is where you manage all your information and engage with your financial plan, and then we have both an iOS and an Android mobile application as well.

We have about four million policyholders, and about a million and a half of them are active users that engage with the mobile app and website.

What made you choose Heap?

The main motivation was self-service. We have a central data team that does centralized analytics through a data warehouse. The team is incredible, but there are only so many of them, and they’re serving the whole company. For more complex apps, they just don’t have the luxury of learning all the details. When we chose Heap we were looking for an analytics solution that the product team could use as power users but that we could also plug in with the larger data team’s ecosystem.

Other analytics tools have never had the level of depth or customizability we get with Heap.

What tools have you tried, and what advantages have you seen with Heap?

At LearnVest, the fintech company Northwestern Mutual acquired, we used Google Analytics and Periscope. GA ended up being too basic and Adobe was too complex and too content-based.

Analytics products just have never had the level of depth or customizability that we’ve been able to get with Heap. Features like Heap’s event visualizer make it easier for product managers to connect the analytics to the specific part of the application that they’re responsible for. The adoption curve for Heap is much quicker.

One of the first magical things we discovered was autocapture. Prior to Heap, we had to be very diligent about how we tracked stuff. It was very easy to forget to tag something, or to forget to call out what we wanted to track. Just the notion that you can set it and forget it has saved our bacon a couple of times.

What kinds of things are people tracking and analyzing?

We do a lot of reports and dashboards. Every team has an adoption report showing the percentage of the user base that engages with specific features. We can even do pro-active stuff like email users that have tried a specific feature or notify users who may have tried a feature but stopped and ask them to re-engage when we make improvements.

We’ve also started more sophisticated analytics like retention cohorts and retention curves. In financial services, customer journeys can be very elaborate. For example, if you want to put more money towards your retirement you and your advisor may decide to increase your IRA contributions. The steps to do that on the website are fairly complex, no matter how well you design the flow. You have to link a bank account. Put in how much money you’re going to add. What’s the frequency? One-time or repeatedly? You may have a tax limit on how much you can put in too. The flow is way more complicated than a typical e-commerce conversion funnel and we’re constantly trying to make it better. Heap’s allowed us to set up longer term goals like OKR objectives, retention cohorts, and detailed analytics journeys so that we can see if our changes are actually driving our outcomes.

We almost take it for granted now that every new feature that goes out has analytics on it so we can quickly see if people use it or not.  That’s something we couldn’t take for granted a couple years ago.

- Joe Pfeiffer, VP of Product

What effects have you noticed with Heap—any big wins?

I think the biggest win is that our product managers are data-driven. The best product managers use Heap — you can tell. They just have more confidence in the way they defend their decisions and explain how people use their products.

Any specific results you can share?

Sure, I’ve got three stories for you.

From 3% adoption to universal acceptance

We had a business stakeholder who was convinced that a particular feature was the best thing ever. A lot of us were skeptical but we built it anyways. A few months later adoption was only 2 or 3%. Without Heap, we would have forgotten to track it, or assumed this loud stakeholder was right and that everybody used it. But instead of saying the stakeholder was wrong, we used Heap to figure out how we could cross-promote the feature in other parts of the app. We kept making changes on it, and close to 100% of users did adopt it. It was such a constructive way to do feature iteration. We’re now much more receptive to new and unexpected ideas, because we have the muscle to study things as opposed to relying on opinions.

A 40% decrease in search errors

When we launched a new internal search engine, Heap told us there were a bunch of searches returning zero results, which is the worst thing a search engine could do. We made tweaks to autosuggest stuff for people that were fat-fingering and mistyping, and built a Heap dashboard to target the decline of zero results. But since we were changing the search engine, there was a good chance that we were going to regress and screw up search results for people that weren’t getting zero results previously. So in parallel, we ran a second report on not removing clicks on the first five results. After we implemented the feature, we saw about a 40% decrease in zero search results.

A 20% increase in engagement

One of the really cool features on our mobile app is the ability to track plans to accomplish big financial goals, like save for retirement if you have student loan debt, or save to buy a house if you’re planning to have a wedding. One of the first steps is to link your bank accounts to your goals, so that we can monitor your progress and nudge you if you need a friendly reminder. This is a key feature that drives engagement, so we launched a feature tour handholding you through linking your accounts to your goals. The tour saw a 20% increase in people engaging. The bad news was a lot of people weren’t even starting the tour, or they would abandon it. The silver lining was that with Heap we knew exactly when the feature tour wasn’t working, so we could actually iterate on that. That helped us break down the bigger problem of use and engagement, and isolate it to a smaller set of things that we can sprint on and iterate on.

Heap is on my shortlist of must-have tools for building web apps.

- Joe Pfeiffer, VP of Product

How have relationships with senior leadership changed?

I feel like conversations are a lot more constructive, because since we are not primarily a tech company, senior leaders may not be familiar with the gory details of the products or the analytics.  But they’re very familiar with the outcomes we’re trying to achieve. So when senior leadership says, “show me what you’ve done,” I can pull up a dashboard and show them metrics that we’ve moved the needle on instead of a laundry list of grab-bag features.

If you had to recommend Heap to a friend, what would you say?

This is going to sound crazy, but I still love this feature in iTunes called “Smart Playlists.” I feel like Heap is similar, in that anything is possible. I can start really simple, but I can get complex with custom rules that say “include this but exclude that,” and “not bigger than this, but not smaller than that.” That type of capability is a power user’s dream. The only other thing that has that level of flexibility is having to learn SQL which not everyone can do.  Heap offers an amazing balance of simple and complex that evolves with you as you become better with data.

PreviousNext

Find your own wins with Heap

Digital insights await! Get started on reaching your business goals today.