Maison Sport knows that independent ski and snowboard instructors often go the extra mile to delight their customers, having worked in the industry themselves! They sought to create a platform that allows users to find their perfect ski or snowboard instructor and book them in advance.
This breaks the mold of the industry, where traditional ski schools hold the power - customers arrive at resorts, book right then and then take the instructor they’re given. For instructors themselves, they are at the mercy of the ski school’s policies - often losing the bulk of the revenue from a booking to the school, rather than the individual delivering the service.
As such, Maison’s platform cuts out the middleman, allowing customers greater choice and control over their booking while saving money, while simultaneously putting more money in the pockets of independent instructors who are self-employed.
Alongside our main objective, we also wanted to:
Acquire new instructors for Maison’s platform
Increase the digital awareness and presence of the brand
Our work with Maison Sport spanned multiple seasons. Initially, we proved the concept of what digital marketing could bring to the table by implementing key retargeting campaigns, capitalising on those users who have browsed their instructor pages or were currently very close to a resort.
The following season, we expanded our work to cold campaigns, using Facebook and Instagram to find Maison new customers across the UK, France, Italy and Switzerland despite a tricky 2020/21 winter season as COVID flared up across Europe.
In the 2021/22 winter season, we took the learnings from that prior season and were tasked with aggressively scaling the account, taking customer campaigns from £2-3k/month in total to £50,000-£100,000 via direct scaling, expanding into new markets and rounding out the retargeting ecosystem. Monthly spend variation was primarily driven by the ebb and flow of travel restrictions as Omicron flared up in December and France dramatically reduced access to UK customers.
Aside from the impressive growth in scaling (and doing so profitably!) our proudest achievement with the client is probably consistently beating their CPAs from Google. Competing directly against high-intent searches with Paid Social, which seeks to interrupt customers who are not already looking for the product is a tall order, especially when tasked with scaling aggressively.
We started by ensuring accurate tracking was in place. Maison Sport has a long sales cycle, as users are often making a booking that ranges from £300-£2000 and will therefore invariably take some time over the decision. We went about setting up advanced tracking throughout the site on Facebook and this evolved through our engagement with the client, with Aggregated Events Management and offline conversions vital to combat the data loss every advertiser experienced on Facebook as iOS14 adoption rates increased rapidly in the summer of 2021.
To get the social side of things underway, we started with retargeting campaigns on the highest-impact potential customers - those who had searched for an instructor and browsed profiles but not yet booked. This allowed us to demonstrate the power of paid social to the client and the results were very strong - we were acquiring customer registrations onto the Maison platform for £35.35 and returning 10.73 ROAS over the season. While overall volume of traffic was low as we went deeper into the funnel, we were able to utilise Reach campaigns to target checkout viewers who hadn’t bought, resulting in 75.57 ROAS.
We also supplemented this traditional retargeting with specific resort retargeting, putting dedicated copy & creatives in front of individuals who had viewed webpages relating to specific resorts. This allowed us to retarget users who were earlier in the funnel in a way that acted more akin to late stage retargeting, where intent and CTRs are much higher, as the content was hyper-relevant to the individual.
This gave the client confidence that their ideal customers were on Facebook & Instagram and so at the end of the ski season, they committed to re-engaging with us for the following season, widening the scope of work to include cold customer campaigns and new instructor acquisition.
This season, our remit was widened to include cold customers. This was great, because we knew the retargeting campaigns were highly effective and profitable and the new cold campaigns would drive much greater traffic into them.
Our remit was to target the UK, France, Italy and Switzerland. We targeted each country with five copy angles, allowing us to establish which messaging was most effective in terms of garnering interest, signups and purchases. We also tested fifteen different creatives - fewer than we would normally test, but sufficient given the lower testing budgets we were working with.
Early on in the season, what we rapidly learned was that effective messaging varied by the market. The UK was happy to try the Maison concept, whereas the French were fiercely proud of ESF, a seemingly national institution when it came to booking ski lessons. Italians engaged fiercely with the platform, showing some of our lowest cost per signup numbers and CTRs, but did not convert effectively to purchases at this time.
As we moved later into the season, COVID spikes occurred and travel restrictions were imposed and resorts were shuttered. As such, we were forced to cut back all spending in the UK, France and Italy as news worsened from November to December 2020. This was particularly disappointing, as interest normally builds to a crescendo at Christmas. We did however take the opportunity to scale the Swiss campaign, as the domestic market was unaffected by travel restrictions and resorts remained open. As such, at a bad time generally, we were able to increase spend fourfold in this market and discover new audiences that ultimately guided us in the next season.
The final part of our work was focused on achieving new instructor signups onto the Maison platform. While they already had a range of instructors, the more they had registered the greater their capacity was to take bookings and the wider their reach across various resorts. As such, we ran lead generation campaigns whereby potential new instructors had to create an account, complete with email and phone number, before submitting an application to join the platform. As such, this was lead generation with a high degree of friction at the front end, which we overcame by ensuring our targeting, copy and creative captured the most engaged users. We tested various audiences to find instructors in the most targeted way possible - testing both lookalike audiences of existing instructors and interest targeting. We also utilised long form copy to sell the benefits of the platform and ensure that the clicks we got from ads were as high-quality as possible. Overall, we added 300 instructors to Maison’s platform in this period.
To wrap up our engagement for the season, we produced a report covering all learnings over the period. This compiled everything we have tested and our takeaways from it (and was additional to our normal weekly reporting), as well as our scaling plan for the next season. This came to almost 6000 words, but was hugely valuable to the client in speaking to their board of directors & inveastors about future plans, and also helped us consolidate everything we had learned and serve as a platform for the next season.
This year, our remit was clear. SCALE. Coming out of a very disappointing season where COVID had cost the client the majority of the season, they were behind where they wanted to be in terms of customer growth and therefore had bold goals for this year from their investors.
Using our previous learnings report, we rapidly built campaigns based on our prior best-performing copy, creative and audiences. Operating at much greater budgets than we had previously, it was vital to put our best foot forward.
This was supported by dedicated copy/creative testing campaigns, to ensure we had a pipeline of fresh, new, well-tested assets. Testing what specific creatives, messaging, copy length and call to actions were delivering results allowed us to optimise spend on the individual ad creatives so that we could maximise our marketing efficiency and the client’s overall results.This separation also allowed us to have “testing” campaigns separate from “Best-Performing” or “Super” campaigns, free of clutter. The best-performing “Super” campaigns could then be scaled aggressively, as well as utilised for audience testing through which we found a pool of new audiences that gave us additional breadth.
For example, in the previous season we had tested various lookalike and Broad (untargeted) audiences, but had found these underperformed relative to a broad, ski interest audience (covering popular ski clothing brands, events, and public figures). To date, this audience has represented more than half of the spend in the main Super campaign, however we were able to scale through additional audiences such as:
What was vital was letting go of our previous assumption “lookalikes don’t work” and recognising we had vastly more data this time for the LLAs to work from. We also were not hung up excessively on the data source, electing to “test it and see”. For example, we logically expected a list of purchasers to create a better quality lookalike audience than people who only signed up for the platform and did not make a purchase. However, the reality is that to date, our Customer Sign Up LLA is outperforming in ROAS terms, at 3.73 compared to the Purchase List at 2.14, as well as providing a cost per signup 40% lower. It was better to test and find this out with a small budget, than purely focusing on “what makes sense” and not testing more broadly.
Once we had these varied data sources, we also tested varied percentage LLAs. This allowed us to assess the impact of broadening the audience - this may dilute the quality of the audience (bad) but lead to a greater pool of potential customers to target (good) as well as lower CPMs (good). In ROAS terms, the 1% LLAs outperformed, but in terms of overall revenue, wider LLAs contributed more strongly as they were able to spend at a greater level in their larger pool of potential customers.
Through a combination of audience testing and direct budget adjustments, supported by constant creative refreshment from the testing campaigns, we were able to scale the Super campaign from an initial daily budget of £500/day to over £2500/day profitably (1.95 ROAS, while also acquiring 274 new customer signups as a byproduct).
All of this traffic naturally helped us to feed our retargeting audiences, which were structured to establish touch points with UK users at every stage of the customer journey:
Combined, these campaigns also recorded 69 new customer signups as a byproduct. This full ecosystem, as you can see from the numbers above, really amplifies the overall profitability of the cold campaigns.
Outside of the UK (the key strategic market) budgets were less but still notable. Similar to the UK, we set up a full ecosystem which supported the overall results, however the Italian market, being closer to the mountains) was dominated by smaller, more frequent purchasers buying for regular weekend trips to the mountains as opposed to a 1-2 week trip with the family complete with flights. As such, purchases were more skewed towards the cold campaigns and we differentiated our retargeting in this market by utilising lifetime budgets to schedule retargeting campaigns on key days. For example, we could hit users who had engaged with the website on Thursdays and Fridays with messaging such as “planning a trip to the mountains this weekend?”, tailoring our messaging and the timing with the context of the local market.
Overall in Italy, this growth in spend radically grew Maison’s market share - almost 100 bookings were made at the midpoint of this season, with 221 new customers registered on the platform too. This is a huge upgrade from our previous testing season (also blighted by COVID) where no purchases were recorded and twenty customers joined the platform.
In France, we knew from prior campaigns the local market was fiercely loyal to the national ESF ski school. As such, we tested new copy angles when we launched the campaign, focusing on handling key objections that came up. These included emphasising the fact that Maison offered private bookings, rather than being lumped into a public group lesson, as well as making clear that Maison’s instructors are qualified to the same level as ESF instructors. This differentiated messaging was hugely impactful, capturing all purchases achieved versus our tried-and-tested angles from other markets, which struggled. We also noted a 98% reduction in negative engagement on the adverts. Previously, French users had criticised/attacked the brand for pushing into ESF’s market, whereas now adverts were handling those objections from the off and actually telling a more positive story, focusing on independent instructors who set their own wages & working hours.
As the season progressed, we were able to scale via tapping into new markets, too. We took the heavily-tested ads from the Super campaign and ran them to strategic countries in mainland Europe, populated with British expats, located in central Europe. This added an additional audience as big as some of our largest UK audiences, adding huge scope for scaling.
We also tested new markets based on the client’s insights on where organic bookings were coming from. For example, if there was an upsurge in Dutch phone numbers coming in as customers booked, we could rapidly launch a campaign into the market based on best-performing assets, translated for localisation and scale spend. Our tracking likely caught 70-80% of direct purchasers (about as good as you can get in a post iOS14 world) but the client monitoring the macro impact in terms of bookings would also feed in and guide budget decisions. This would allow us to scale or pull back, monitor the impact in both the ad account and what the client sees, and assess results. This collaborative relationship was huge in enabling us to maximise results for the client in a world where tracking is less effective.
Heavily localised targeting was also used effectively, taking advantage of Facebook’s ability to geotarget very accurately. We were able to create audiences of individuals in close proximity to resorts and target them with adverts. On top of this geographic targeting, we could layer additional targeting to narrow the audience as desired however we found that the narrow geographic targeting was sufficient.
We also added an additional layer to the retargeting ecosystem to supplement the variety of new cold campaigns we were running. In this, we sought to retarget past users to drive re-engagement with the platform.