Develop a competitive advantage by exploiting site search behavior

Business

Urban Sports Club offers a monthly subscription that gives members unlimited access to its network of 6,000 partner gyms in 50 different sports across Europe.

Requirements

Urban Sports Club members take out memberships to access all the network's sports clubs at lower cost, as well as specific sports activities.

To remain the leader, Urban Sports Club needs to sign up new partners faster than the competition and offer a relevant product range.

Business development teams therefore need to identify the partners or types of sports by country, city, season, etc. that generate the most interest among prospects.

Objectives

Analyze the search behaviors of site visitors to identify the types of popular searches that Urban Sports Club doesn't seem to offer enough of, such as partners, sports, cities or countries. But also to identify the searches that most encourage users to subscribe.

• Do some cities have a high volume of searches but a less relevant offer?

• Are certain types of sport more popular in certain cities?

• Are there any particular sports clubs that are in demand but don't currently have a corresponding offer?

• What is the ideal or sufficient number of partners in the offer to convince prospects in certain cities?

Proposed solution

• Establish a definition of a successful search for the user (ratio select_item / search) which gives us a % of search success. That is, the % of times a sports club is selected in a list of search results.

• Si le taux de success search < x% pour un sport dans une ville ou quartier, alors on en déduit qu’il n’y a pas l’offre adaptée. Les équipes de Partner Management peuvent ajuster leurs stratégies de démarche commerciale des partenaires en fonction.

• Measuring search behavior : search, filter, spell , select_item

• Define search attributes to understand performance based on certain attributes.

Implementation

Architecture

• Update data collection in the data layer

• Data collection in GA4

• Data export and reprocessing in BigQuery

• Creation of a tool in Looker Studio to explore insights

Example of updating the data layer to collect the right data :

dataLayer.push({
  event: 'search',
  search: {
    plan: 'm',
    location: {
      country: 'germany',
      city: 'berlin',
      district: 'kreuzberg'
    },
    activity: {
      category: 'boxe',
      partner: 'exfighter'
    },
    results: {
      count: '89',
      view: 'map'
    },
  },
});

Project duration

Overall, the project took two weeks to implement. The development team was dedicated, but there were many recipe points to check and some back and forth with the development team.

Results

For the Product (Website) team and Partner Managers :

• New KPI for the site: search success rate

• A tool in Looker Studio to identify opportunities

• Enriched data collection for the BI team

Further information

Here are a few ideas on how to take this project even further:

1. Anticipate market demand: integrate Google Trends search terms by city / sport / time of year etc. to cross-reference localized trend searches with the proposed offer.

2. Reduce acquisition costs: Create retargeting audiences based on clusters of identified users and their search behavior.

3. Enrich customer knowledge: Identify studio combinations or types of sports frequently searched for together.

4. Track partner performance: Once partners have been integrated into the offer, track the performance of each partner to assess their attractiveness and contribution to conversions.

5. Test different search result models: thanks to server-side A/B testing (to test different algorithms serving search results) and the search success KPI.