Case Study

Merging the HomeAdvisor and Handy business models in order take up more space in the online home services market.

Quick overview

In late 2018, HomeAdvisor acquired Handy when they saw an opportunity to combine our business models and take up more space in the home services market. Before the acquisition, we knew that Handy’s online booking model was not the traditional way to buy home services. Our model worked really well for simple, commoditized services that people believe can be priced automatically, like home cleaning or TV mounting. It did not work well for bigger or more complicated projects, like painting or plumbing jobs.
‍
Luckily, through the acquisition, we gained access to HomeAdvisor's tools, data, and expertise in the more traditional home services business model. HomeAdvisor’s business model caters to users who want to speak to a professional on the phone before booking and to users who prefer not to pay for home services upfront (needs Handy has never been able to fill).

The problem to solve

Post-acquisition, the question for our product team to answer was pretty simple:​

How can we offer Handy users access to HomeAdvisor professionals and services without cannibalizing the existing Handy business?

HomeAdvisor generates revenue when users contact companies or professionals via the HomeAdvisor platform. We knew that if we could expand the reach of those pros by exposing them on the Handy platform, we could start generating additional revenue right away. However, we didn't want to detract from our existing business and cannabalize our own model by distracting or confusing users who wanted to book directly with us.
Existing user flows before the acquision.

Solution

Integrate HomeAdvisor’s more traditional business model into Handy’s online booking model.

Our overarching objective was to make Handy useful for more people by integrating HomeAdvisor’s traditional business model with Handy’s existing booking model. We broke this out into more specific goals to align around clear measurements for success.

From previous user interviews and testing, we knew that most new users expect to see a list of professionals in our booking flow and many people are put off by the idea of paying for a service upfront.
‍
With the HomeAdvisor acquisition, we gained access to a base of local professionals and small businesses with whom our customers could speak to on the phone, discuss project details, and negotiate pricing.
‍
By combining both fulfillment options (Handy's "Book Now" and HomeAdvisor "Request a Quote") we would be able to serve users who wanted to book services instantly online (existing Handy customers), as well those previously unserved customers who wanted to go the more traditional route of talking on the phone and getting pricing from multiple local companies.

Competitor research

Understanding user expecations

Once our goals were established, I started working on mapping out how this new world of multiple service fulfillment options could look on the existing Handy site.

The first thing I did was review how other companies were showing multiple service fulfillment options in their user flows. Our main comps for this came from Yelp and Thumbtack, since we felt like they were doing the most interesting things in this space and mapping closely to where we wanted to be.
User flow I created from my shadowing sessions.

Sketching

Brainstorming solutions

A few members of the product and marketing teams met several times to brainstorm solutions. We talked about what theHandy/HomeAdvisor merger could look like in a successful, future state where users could come to our site and decide if they wanted to book with Handy or connect with a local company through HomeAdvisor's API. Thinking through options as a group didn’t define the solution, but it helped us decide what types of features and changes we wanted to test into and where they could live.
Some of the sketches that came out of our brainstorm sessions.

Defining a solution

Seeing an opportunity on the payment page

The next step was thinking about our existing product and noting the biggest areas of opportunity. Currently, almost 90% of the users who click on a service end up dropping out of the flow once they get to the payment page (the first time we show pricing)without making a booking. This felt like the right place to start testing other service fulfillment options.

Setting Goals

We decided on two design goals and two high level KPIs that, if accomplished, would result in a successful product launch.

Our design goals were to increase user choice and make the product more useful and engaging. Our plan was to better display available information and service fulfillment options.

Our product team worked on the business KPIs. When users request a quote on their project from a HomeAdvisor company, we monetize on that lead, so net revenue was a clean metric to measure whether our changes increased the number of quotes requested without cannibalizing the existing net rev from fixed-price Handy bookings. In order to make sure we were hitting the second part of the design goals (“make it useful for more people”) we added conversion tracking.

Design Goals

We decided on three design goals that, if accomplished, we believed would result in a successful product launch.
1. Increase User Choice
Better aggregate available information and service fulfillment options to create better choices.
2. Enable HomeAdvisor Pros
Provide HomeAdvisor pros with a new platform to acquire customers.
3. Increase Engagement
Get the 90% of people who don’t book with Handy currently to take another action.

Metrics Tracking and KPIs

When users request aquote on their project from a HomeAdvisor company, we monetize on thatlead, sonet revenuewas a clean metric to measure whether our changesincreased the number of quotes requested without cannibalizing the existing net rev from fixed-price Handy bookings.However, in order to make sure we were hitting the second part of the goal(“Make Handy useful for more people”) we adding conversion tracking.
1. Overall Net Revenue
Our goal was to grow the number of quotes requested from HomeAdvisor pros without cannibalizing the existing net rev from fixed-price Handy bookings.
2. Funnel Conversion
There are a ton of ways to increase net rev without creating a more useful product. Adding conversion tracking helped us determine if we were really creating a tool that served more user needs.

New User Flow

Mapping the MVP

After considering the ideas from our brainstorm, the user flows of our competitors and the flow of the existing Handy product, we mapped out our new user flow.

The biggest change was the addition of a "Booking Options" page (step 3 in the figure below) where we would prompt users to decide if they would book with Handy at a fixed price or request quotes from local professionals (a list that would be powered by the HomeAdvisor API).
The basic MVP user journey map our team decided on based on research and existing UX.

Wires

Designing the new service fulfillment options page

At this point, our team agreed that we needed to present the option to Request a Quote and the option to Book with Handy on the same page (step 3 in the figure above). However, we didn't have a clear vision for the UX/UI of this page. 
‍
Based on our competitors, it made sense to show a list view with Handy as one booking option and the HomeAdvisor companies as other options. However, we were really nervous that showing the HomeAdvisor companies as an equally weighted options would cannibalize Handy bookings. ​
We mapped out the different display options that made sense.

A quick A/B test in prod

In A/B testing, the "middle of the road" option was unsuccessful

To get an initial read, we tested a version of option B (Banner with a CTA) on our existing payment page. We chose this option because it was a relatively small lift for engineering and we thought it would give us a good idea of how adding another option to the page would affect conversion, as well as indicating user interest. 
‍
Ultimately, however, the banner got really low engagement and did not increase or decrease net revenue, so we ended up killing the test. Through user testing, we learned that most users were skipping right past the banner. The users that did see the banner found it confusing. They said it looked kind of like an ad and there was no clarity around what would happen if they did hit the CTA. 

User testing more ideas

Running usability tests on the more extreme design option

After our initial A/B test was unsuccessful, our team decided that based on user feedback, we needed to go with the more extreme version of the UI and show users a list of booking options. I used Invision to build prototypes of the flow above so we could user test many different versions of the list UI and see which one made the most sense to people.
We mapped out the different display options that made sense.

Key learnings

We ran over 60 usability tests on 4 different designs

The most valuable thing about testing with so many people was the clear pattern that emerged between "Book Now" users and "Request a Quote" users. Even though user testers were not real customers, they tended to feel very strongly about choosing one booking option over another. We asked them to describe their thought process and decision making and we were able to use those clips to create general user types for the team to reference.

Overall, the different designs we tested were pretty similar in terms of user flow and information provided on the list view. We were looking for small clues to tell us how people would perceive the options and how we could design it to optimize for clarity and decrease decision paralysis. ​

We were also working with a finite amount of information about the HomeAdvisor companies because of the way their API was initially built. The user testing helped us learn what type of information users were interested in, so we knew which API changes to ask for from their team. 
We mapped out the different display options that made sense.

Initial product launch

A/B testing the MVP on smaller services

We rolled out an MVP version of this page as an A/B test to 50% of logged out users on some of our smaller Handyman services. We launched with the key features we felt were central the the success of the product, like basic company information and success landing page for Request a Quote. 

​Based on our user tests, we knew we were missing key features like customer reviews, but we opted to release earlier so we could start getting data back as soon as possible. 
Screen shots from the MVP user flow.

Initial data

A/B testing this flow across multiple handyman services allowed our team to get data back quickly and determine if we were going in the right direction. Our initial response was very positive, we were able to hold Handy conversion in line with the control while growing overall net revenue through the quote request flow.
Total net rev/session +126.2%
Net Revenue / Session: $6.91 in Test Variant vs. $3.05 in Control
Total page conversion 14.6%
Handy conversion at ~6% and Request a Quote Conversion at ~8%
Handy payment to booking completed remains flat
Handy conversion was flat at ~6% across both variant and control

Improvements & iterations

Since our initial A/B test of this flow showed such positive results, we were able to invest more engineering time to make the V1 updates. We knew from the user testing that reviews were most important thing to do add. We designed the reviews view as a scrollable modal, which included a sticky CTA for the mobile web view. ​

We also made some small usability tweaks to the list, most notably by adding the cost guide, removing the section subheaders and adding small helper text under the Request a Quote CTAs. We decided to remove the subheaders after getting user feedback that they were more confusing than helpful.  This also helped us work towards a future state for this list where the options are dynamic and we can push the best option to the top.​

The last thing we did was add the ability to request additional quotes to the Success page. This helped increase the number of total quote requests while ensuring users get the best pricing on their project.

Iterations & improvements to the flow that ultimately launched to all users.

Thanks for reading!

View Next Project

It's not the end,
it's just the footer.

Some ways to reach me:

Let's stay in touch.
© Molly Jaffe 2023