Gumtree Pets Project
(Case study writing in progress) A proof-of-concept project to test how well multiple multi-disciplinary product / design / development squads could coordinate on a sweeping initiative at speed, while also utilising new user mindset research and quick fire insights to build features for a user set they had yet to monetise.
THE CONTEXT & OBJECTIVES
Search and Trust
During 2024, the biggest themes coming out of customer frustrations that were seen as directly impacting the business (outside of parity of features with competitors which were already being worked on) were:
- Search: Users had difficulty finding what they were looking for due to Gumtree’s awkward taxonomy, lack of granular search result filters, and items / posts being placed in the wrong category within the posting flow (re awkward taxonomy). This impacted engagement metrics: buyers couldn’t find what they needed (easily) and sellers’ weren’t connecting with the right customers.
- Lack of trust of Gumtree on those who used it: Many online marketplaces were struggling with scamming on their platforms and despite a mountain of work done to reduce scams on site, users still wanted more reassurance that they could trust the sellers – they wanted more info, more reviews and ratings, more verification… This severely limited who used the platform to buy and who was willing to sell on it (impacting audience engagement and revenue from sellers posting ads).
How to start addressing all this?
Launching a massive project with a contained deadline of 4.5 months that served to be a proof of concept for three distinct aspects of Gumtree’s approach to product:
- To finally dig into the spaghetti-like back-end code and create better, faster, more consistent ways to add attributes and filters to the search journeys, helping users find what they are looking for more quickly.
- Utilise recent research around user mindsets to help surface information in ads and about sellers to instill better trust with Gumtree and the buyers and sellers on the platform.
- Build an experience and feature offering framework that could allow Gumtree to monetise a different tier of sellers, making Gumtree a more attractive platform to sell or post compared to their most immediate competitors.
THE SET UP
- My role: Director of Design and Research / Project Lead
- Company: Gumtree
- The Team: Head of Product, Head of Delivery, Head of Engineering, designers, researchers, delivery managers, front end and backend developers, commercial team
Prior to this project, features and flows for the Gumtree site and apps were implemented in an Agile-like manner through multi-disciplinary product / design / dev squads separated out by specific user areas:
- Buyers / searchers
- Sellers / posters
- Accounts and communication areas both sets used
Due to the objectives and scale, though these squads still technically worked autonomously, the number of dependencies and knock-on effects of features and tech being implemented meant all of these squads had to work more collaboratively and communicate more frequently on this project than other work they had done in the past.
Yes, why start with Pets?
Since easing of Covid lockdown, coupled with some unsuccessful pricing changes pushed to sellers, plus all of the rising competition online, the Pets vertical had been in decline. It was seen as the one vertical, wholly owned within Gumtree’s tech stack, that was the least risky (due to it’s overall revenue impact) and had the most to gain from any improvements to both search and trust initiatives. So off we started.
THE PROCESS
My role for this project
De-facto Programme Lead (and UX Lead), I was responsible for coordinating ALL the squads to focus on the same outcomes. The bulk of my role, outside of my day to day work around research strategy and user-centred design direction, was to kick off this massive initiatives and set up structures, ceremonies, and processes to help ensure all squads were communicating adequately and had visibility of research, design and dev work in progress in all squads to prompt collaboration, refine sequencing of build / release and ensure the delivered experiences met the larger business and user objectives.
Getting things started
We knew the broad strokes of what needed to be done in terms of broad objectives, but before we started to design or code anything, we needed to clarify the opportunities and insights in the Pets space specifically; trying to tailor for the vertical while identifying areas that could go across most other verticals. We started with an epic brainstorm with those in Product & Tech, Commercial, Customer Service and beyond to get all of the possibilities on the table. Next came the evaluation of the research and insights we had already.
Mindsets for Pet sellers and buyers
The User Research team had just finished a massive project on User Mindsets, something more flexible and expandable compared to our initial user personas. The focused mindsets around Pets helped everyone concentrate on the specific needs of Pets users, especially breeders advertising on the platform.
We have direction, now the details
The mindsets help set a direction to further discovery to be done to help prioritise which aspects of the taxonomy (aka “attributes”) should be implemented first as per what users found most useful and important, and which attributes helped to instill more trust in the seller / breeder and Gumtree as a platform to safely buy and sell pets.
A new audience emerges
This research also helped to support an internal idea of a new market to enagage and eventually monetise: smaller, professional breeders for cats and dogs.
By focusing on smaller breeders, we could:
- Look at how to build trust in our pet sellers (giving more visual prominence and priority to breeders with certification)
- We could more easily normalise and inforce PAAG requirement pet listings
- We could look at a larger set of features and support to offer smaller businesses across multiple verticals to help separate them out as professionals actively trading on the platform vs occasional private sellers.
Final planning before go time
We now knew the priorities of the user and tech had come back with all of the attributes that already existed in the database but had yet to be surfaced in the taxonomy or the UI. Great.
But every squad depended on at least one other to get any of this going – some attributes (e.g. breed) couldn’t be shown to the prospective buyer searching for a put until a seller had added it to their advert. The seller couldn’t add the attribute to the advert in any way for the system to pick up as an attribute until the search team added that attribute to the database (or made what was already there accessible to the UI). The squads needed to know how their work impacted another team.
Dependency and Sequencing: I led a workshop with all members of the squads working on this project. This helped clarify the prioirities and sequencing of work in the Product Roadmap.
Ceremonies and Oversight: From there I also put together a list of ceremonies to ensure communication and visiblity between squads, processes to make sure Trust & Safety, Commerical and Customer Service were involved at appropriate points and steering groups to ensure leadership could give feedback and give context for any shifts in priorities.
RACI: Most importantly, I worked with the Product, Delivery and Tech teams to define a RACI for decision making as these groups had not worked this integrally for a while.
ROLLING IT OUT
The starting pistol fired and we were off to the races of design and implementation. Most of the heavy lifting for what got implemented within this timeframe happened on the development side of things. The prioritisation of which aspects to tackle that would show up in the user interface was heavily influenced by research findings and work done in conjunction with the design and trust and safety team.
Here are some samples of work that did get launched in the interface to help improve audience engagement and revenue impact (from sellers within Pets).
More project images coming soon
RESULTS
It was a very successful proof of concept.
The work produced 3 distinct traunches of output:
1. Knowledge, output and a tech playbook for how to add or adjust attributes into the taxonomy to aid in delivery better search results for buyers (more relevant ads) and better ad categorisation for sellers (making sure their ad shows up where buyers expect to find them). This laid important groundwork for larger, future taxonomy work.
2. How to integrate larger strategic research along with quicker turnaround user insights into a fast moving, complex project. This led to more effective, user-centric prioritisation of features which had immediate positive impacts on user engagement, sentiment and revenue.
3. Clear insights on how to approach a commercial package of features and benefits for smaller businesses trading on Gumtree.
Specific outcomes
• Placing an ad for pets became more specific to the vertical. Feedback from pet sellers and breeders was much more positive.
• In launching of dog breeds within the taxonomy, all dog related ads had breeds attached to them by week 3 after launch. Breeds in the cats category launched soon after with similar results.
• Verifying breeders – by allowing cat and dog breeders to be verified, this opened the door to having more prominence in the search results and badging to help build trust with buyers / adopters.
• Lower the listing fee for pets, based off of extensive research with breeders (as they post most often in pets), saw an uptick of 16% in placed pets ads within the first month.
Want more detail?
This is just the topline . I am happy to go into more relevant detail. Just contact me and let me know which areas you’d like to know more about.
