Case Study: Growth Strategy for a Food-Tech Delivery Business
While fintech may still have the leading spotlight among entrepreneurs and investors, food-tech has gradually become an important part of the startup world in recent years. The success of Deliveroo in the UK and abroad has prompted incumbents such as Just Eat into extensive marketing and significant M&A activity. The buzz is felt around the world with each region having its dominant players who are keen to expand (or keep expanding) internationally — Germany’s stars include Foodora, Foodpanda and Delivery Hero, India has its Swiggy and Zomato, and USA has the likes of GrubHub and DoorDash. Uber Eats is not the only example of a major company to engage in food and technology. The prospect of autonomous delivery and the untapped potential therewith, has led to Tencent’s investment in the robotics startup Marble, and seed funding for delivery business BoxBot from Toyota among others.
This article takes a stab at framing the growth strategy for a startup in this realm of growing interest, and briefly comments on some of the relevant opportunities and challenges.
In order to lay out the problem in a structured manner, Figure 1 shows a simple but (hopefully) effective framework for a starting point. Winning in the long term could involve increasing your margins via either reducing cost, for example by identifying synergies, or boosting revenues. The more exciting latter path (which I have limited the scope to) could be divided into three segments. One way is to extract more value out of the current business as is, for instance by renegotiating commission rates with the partnering restaurants, or by focusing marketing efforts on higher-end meals. The second lever is vertical integration that involves entering other parts of the value chain, for example, also making the food that you deliver. A key example of such an initiative is Deliveroo Editions, kitchen hubs where featured items from upscale menus of Michelin star restaurants are made for delivery. Thirdly, horizontal integration could be achieved by expanding to new geographies, increasing size inorganically by going down the M&A route, or as we dig deeper in the rest of the article, by offering new products or services.
An intuitive way to bucket the many possibilities for horizontal expansion is to apply a customer lens to new business offerings, which renders three main categories as depicted in Figure 2. The diagram continues to show a brainstorm of a few subcategories. For example, fresh products could be added to the delivery range such as raw fruits and vegetables from local outlets, or a subscription to daily milk delivery. Another would be to deliver hot drinks from artisan cafes to large offices during working hours. Foodsby is an example of a startup that has gone for the professional niche, supplying lunch orders to office buildings.
Having a retail presence in the form of packaged ready-meals (like by Boots), branded food condiments (similar to Nando’s sauces), or off-the-shelf food kits (think of Gousto or Hello Fresh in a grocery store) would yet be another possibility, though this would be a substantial shift from the lean delivery business into bulk manufacturing and distribution, and therefore the benefit from any potential adjacencies would be scarce.
Directly leveraging the output of the current business, be it experience, expertise or (most importantly) contacts, is another key method of expanding one’s business into new markets. A food-tech business could use its established network of restaurants to form a marketplace for kitchen staff, especially for sickness or holiday cover. A major proportion of restaurant-related jobs such as dishwashing and table waiting are taken up on a zero-hour contract basis, and would therefore be ideal candidates to include on the extended platform. Culinary contacts would also come in handy for setting up social cooking sessions to professional master-classes or even accredited courses.
In the service range category, another natural extension of the core business would be to offer catering for small to mid-sized business or private events (as done by Pret Delivers). The unique value proposition would be an aggregated service from multiple suppliers in a single order, hence resulting in a lot more choice and customisation for the user when compared to that offered by a generic catering company.
As with all businesses especially those in the digital domain, the key question always revolves around enhanced customer experience. One approach in our context is to look for possible tweaks or add-ons to the core service. For the health-conscious, this could involve joining forces with popular calorie-counters or macro-trackers such as MyFitnessPal, or make proprietary ones (recorders of your five a day maybe?). For city workers, this could be easy-to-order advance ordering for the entire week in one go, pre-set menus for such orders, or smart reminders to order on time.
There is also an enormous scope of using artificial intelligence (AI) in the food-tech industry. Delivery by autonomous robots or self-driving cars has been the talk of the town recently, with promising pilots seen from many startups including Postmates, Waymo, Kiwi Campus and Nuro. It seems we may not be far away from witnessing this as a norm across the industry, as developments in computer vision and deep neural networks are scaled up to meet commercial demands. However, the potential of AI extends beyond robots bringing you that takeaway from Patty & Bun. A recommendation system based on past orders and individuals similar to the user could be set up in a standard collaborative filtering problem. Enhancements could be made through deep learning to offer a two-level personalisation, such that not only is a restaurant recommended, but also specific items from its menu. This could in turn be fed back into the previous proposal of (semi-) automated ordering in a city-worker type subscription model. Similar to Spotify, suggestions for new tastes to excite the palate, or information on trending items and habits could be added. However, as said by Andrew Moore from Google in a recent interview, AI is not some magic dust to be sprinkled on a business — it requires a substantial resource of time, money and expertise, for all stages from problem definition to solution design iterations to evaluation and testing, before anything is rolled out.
The long list of proposed ideas would then need a deeper dive, employing initially a high-level assessment model. Figure 3 demonstrates a slice of such a model that comprises of key considerations like defining the business model and the associated value proposition, an overview of the current competitive landscape and any barriers to entry, along with adjacency to current business or potential cannibalization of it, along with other main challenges. All of this could then be summarised into two key dimensions of relative feasibility and value, which could subsequently be used as the guiding criterion for shortlisting the initiatives for further concentration.
In the practical world, this value versus feasibility study is no trivial task, and may need a significant investment, a wide outlook, and an essential depth of thought to be properly carried out. The even more strenuous task in such strategy exercises is drafting the long list of ideas to consider in the first place. To demonstrate these immense possibilities, I would leave you with Figure 4 that shows a few more ideas for exploration, plus a mention of the American startup Dutchie that is an online platform for ordering weed, and is backed by Snoop Dogg!
This is original work carried out in my spare time. Please let me know of any errors, omissions or improvements.
