Dario di Feliciantonio on scaling acquisition for the largest bartending school in the world

As an Italian I must know how to make a Negroni. The whole bartending culture appealed to me even before I joined, because it’s fun, uniting and it’s global. I knew it would be a fascinating place to work when I came for the interview to the office that looks like a bar. We sat on barstools and spoke on the counter, with all the bottles behind us. That was the moment when I realised that the origin of the business will make the job entertaining. The business is thriving and celebrates 70,000 graduates, despite being hardly hit by Covid.

Analyzing all data about our past and current students has been crucial to understand their perfect journey. Some interesting conclusions that we have analyzed is that our students’ demographics range across 18-30 age and are mostly men. In Southern Europe for example we have two different buyer personas. The youngest ones from 18 to 21 years old in major part are high school recent graduates that are looking to shortcut university path and quickly enter a fun labor market such as the bartending industry. The second buyer persona consists of older people looking for a different professional challenge and trying to completely detour their career from the job they currently have. We also have a 3rd buyer persona who is made of youngsters from all ages looking for a short and fun experience where they can meet new multicultural people, acquire new skills and spend time traveling across sunny places. Their future in bartending is an option but probably not a goal that is set in stone. In all these 3 cases, most of the time these potential customers do not know that bartending may be in their professional future and what do we offer. Given that, we play with targeting broad audiences to stimulate an unaware population that bartending can be the next professional bet. Another trick is interest-targeting for people already in the business looking to complete their CV. The third is the audience willing just to embrace the bartending world. Those we can find by creating lookalike audiences of our most engaged graduates.

The part where I can have the most fun and control is obviously at the creative level. We try to leverage specific concepts in our marketing messaging and shuffle our USPs with different buyer personas to streamline learning. We’ve tried vague and exact explanations of what the course is about and generally, creatives that set clear expectations are the ones that at this stage give best results. Over time we have learned to separate creative testing and other kinds of experiments such as campaign types and audiences. We upload our creatives, for a multivariate testing across all cities, which lets ad platforms define the winners and deliver content based on its own methodology. Then we handpick winning assets to the performance campaigns and keep A/B testing them in specific cities to find those that could deliver longer. We also benefit from being a wide bartending ecosystem. We have three tools at our disposal: first, our IBC Global Certification that is recognized all over the world. Then Matchstaff, our employment portal where we notify former students of potential job opportunities globally. And finally our school network is distributed globally from NYC to Sydney, Goa and Phuket where anyone can take our courses from 5 to 8 different languages. The only case we do things in house is if nothing is on the market or the solutions are not customisable enough. But with the current explosion of SaaS businesses it doesn’t make sense to think of having everything in house anymore. The other way around is true too: you always need a part of internal logic and thinking on top of the tools you use.

We have more than 25 schools spread out across the world though we do not limit to target potential students only in those specific markets. What’s important is that our buyer is open to travel for an experience. For example we have 2 schools in Mexico and Chile but those are key references for all potential students in the whole region, from Colombia to Argentina and Brazil. The same way that Barcelona and Mallorca are great attraction schools for German, British and even American future bartenders. If we look closer, our best-seller course is worth approximately 2000 euro and it’s not something that usually converts in a day. The conversion path from awareness to conversion is taking from three to six month. We leverage top funnel actions through digital marketing and bottom funnel efforts with our international team of in-house sellers. How we work is, we start from an hypothesis that we benchmark and match it with our historical results. That guides us to the acquisition hypothesis that we put on continuous testing. Even though our business was fiercely hit by the pandemic, we were able to increase our testing budget and at the moment 30% of our advertising budget is addressed into experiments. To scale our learnings from experiments across markets, we keep them in tiers – similar markets that have shared behavioural and purchasing patterns. Successful experiments are being applied to other markets belonging in the same tier. For example, the UK and Latam are in different tiers that need appropriate tailored testing. But, if something worked in Ireland, we scale it to Canada and the UK. Such an approach helps us to balance between scale and granular learnings, without wasting too much budgets on similar markets.

Historical data helps us to catch seasonality and learn about general trends. We look for patterns to understand which destination is preferable for students from different cities and run customer interviews to define why. As the consideration journey is long, I don’t rely on location visitor targeting. If I know that the moment of Germans planning to visit Mallorca is approaching, I A/B test driving them to the landing page with a course in Mallorca versus the general page that lists all cities. Over the years some patterns became inviolable, like Americans always interested in taking a course in Mexico and vice versa.

The reality is that we have 25+ schools around 5 continents, each with their own local social media profiles, alongside the global accounts we manage from the central marketing team… so all in all as a brand we manage a total of 58 separate profiles on social media. To manage everything, last year we kicked off a new centralised approach to social media based around a Social Media Management System (Agorapulse) and defined our new digital content strategy. We have brand guidelines and playbooks, but our 25+ schools worldwide all run their own social media profiles independently. We love this local approach because it lets the individual style and character of each school and destination shine through. That said, as a global brand we need some form of governance and control, so we centralised all profiles in Agorapulse giving global reporting, social listening, content scheduling and centralised community management across all 58 profiles. For example, we can now push a post live across all 58 profiles with a few clicks. While content production & posting for school profiles is done locally, community management is done centrally. Giving first-class customer service on social media is a big part of our brand, plus we wanted to take the strain off our schools and let them focus on the courses & students rather than answering messages on social media.

Facebook and Snapchat are our best performers in capturing top-funnel demand. As potential students visit the website, our goal is to turn their interest into a passion by offering a free course. No matter if you plan to become a bartender, or just want to amaze your friends with new shaking skills, a free course is an entry to this world. We then use automated email marketing and this flow returns the lowest CAC. The engagement on the course and in the funnel impacts the lead qualification, a thing that helped us to unite product, sales and marketing teams. Before this, all teams worked in silos. What we’ve experienced is that without cooperation and uniting metrics, each team could add fuel to their own metrics and KPIs, disregarding that each one depends on the output of the other. Working with such structural silos causes a low-quality acquisition, affects retention and correlates to a negative output in revenue. Ironically there can be cases where each department even reached its own goal. The final outcome still ended up being far from perfect.

Attribution can be painful. Given this long journey, if a booking happened without any interaction from the sales team, we give full credit to the last paid click interaction. In case the seller interaction happens, we work backwards to detect the healthiest channel to drive qualified users. From time to time we try different approaches, for example during promotions, we try to go directly straight to the bones and try to drive sales without account managers. If there’s no sales interaction happening, we warmup leads in the funnel and attribute sales based on the last click, as long as we have it. On top of that students are often asked how they knew about the school in their first training days but these are only used more as observational data to contrast with our acquired information. We have an ambitious roadmap and we plan to move into a full data-driven attribution by the end of the year, combined with periodic conversion lift studies to authenticate the results. LTV hasn’t been a KPI we rely too much on, our product is designed to offer our students the best bartending training in our best seller 4-week course and set them up to a professional career in all venues. It is true that some of them come back to complete their CV with more advanced courses but this is not our main focus at the moment, so one of the main challenges is that our students’ turn-over needs to be quick and dynamic. CAC on the contrary is an important KPI. We adopt a thorough strategy integrating both organic and paid drivers with a high focus on content and a final human interaction.

It gets really complex as although we are in the education vertical we are often perceived as sponsoring alcohol consumption and walk a delicate line in the advertising policy. Some markets are more restrictive than others, while in Europe we can easily target anyone above 18, it gets to 21+ in the US and 25+ in Sweden. For instance in India it varies depending on the region. Generally speaking we produce two sort of creatives, in one we may have captions of bartenders engaged in the art of making cocktails, for other we markets we have a more conservative approach, hiding the drink itself or focusing on different USPs such as the certification, the job opportunity, our worldwide destinations and the hands-on practice that you can get in 4-week. Other than that, when you operate globally as we do you have to find a good tradeoff between customization and scalability and also reduce human effort as much as you can. We do not tend to change our approach from market to market except where it’s required, though we always try to localize a sample of our creatives for a specific market, through different elements whether pricing, people, location and specific details to tailor the user experience.

I’d like to give more power to small businesses to compete for the attention of potentially ideal customers with the big corporations. For years and years it was an unfair game because consents weren’t managed properly and ad consumers had no control over their data. Now that it’s changing and ad management is becoming a more transparent process I see a possibility for customers to give more advertising power to brands and products they like. This could shift the current balance of powers. Imagine giving local brands and more entertaining and valuable ads the right to be shown more often to you, no matter what their budget is. Data-responsible advertising could bring this to life.