Growth: What I learned from building growth engines for Marketplaces — A play of engines (Part I)
Marketplace, as we all know, is a dynamic interaction of demand and supply. In this article, we will discuss how to develop an efficient demand growth engine — which is an interaction of multiple engines such as acquisition engines, conversion, incentive, and pricing engines.
Many times, when we talk about strategy, we assume that saying ‘increase conversion by 10%’ is a strategy. It is not. That is a goal and I am a big believer in not just setting a goal but setting up a system — a perpetual, self-reinforcing system that fuels growth. From my own mistakes, I have learned that:
We do not rise to the level of our goals. We fall to the level of our systems — Atomic Habits
So in this article, we will not talk about setting goals but setting up a system. Here are a few objectives that I would like to achieve with this set of articles:
- Discuss what it takes to set up a system that helps create a systematic, deterministic, sustainable, repeatable growth engine.
- Talk about various components of a growth engine.
- We will also see how setting a growth engine is not a siloed function but a highly interdependent cross-functional concerted effort and why companies that fail to bring Product, marketing, design/UX, Analytics, Pricing, Engineering, and Data Science under one team that march together fail to create an efficient marketplace.
- How inefficiency in one system leads to self propounding inefficiencies in other parts of the system.
- Finally, how to identify and develop a strategy that slowly (Sorry I am yet to learn a shortcut for sustainable growth) brings the growth engine into motion.
With so much to cover, let us get started. This is the blueprint of what we will go through in this series:
First, let us look at the various components of a growth engine and discuss the function of each component.
Acquisition Engine — In a strict sense of demand and supply, we can also call this acquisition engine. But in the diagram below, I used the acquisition engine to depict all acquisition channels. The objective of acquisition channels is to bring ‘high intent’ traffic to the funnel. The two channels that all of us might have used for user acquisition would be the suite of FB channels — FB, Instagram etc. and Google Search and its network.
The objective of the acquisition engine is to come up with a game plan to find ‘high intent’ users — a perennial problem of marketing. Typically measured through CAC & RoAS
Conversion Engine — The next engine is the conversion engine, my favorite because building this requires the team to be highly iterative, experimental, and comfortable with the arts of user psychology and science of product analytics. The Conversion engine or Conversion Rate Optimization Engine (CRO) can typically be best understood by the funnel:
Let us dissect each section of the funnel — time to dive deeper
- Awareness: The first part of the funnel is called ‘Awareness’. This is the surface area where marketing efforts or other efforts touch the user for the first time. Marketing channels will bring the users to one of the pages on the product. This is typically the home page but more and more new landing pages customized for a purpose are launched to explain the value proposition to the users — there is a risk to customizing pages for various channels and segments but that is for another time.
- Search: Users fall into the product funnel and get to the search page. With the supply engine working at its full steam underneath, an efficient marketplace is all about the density of quality supply. However, dense supply also presents a challenge, how to show supply in an ordered format to the user — the ordering should present the most relevant inventory to the user first. This is a huge challenge and displaying a list of inventory requires a deeper understanding of user psychology, discrete choice analysis. Despite all of this, it is still an extraordinary cognitive load on users — leading to decision paralysis, decision fatigue, and drop-offs. If you are not able to understand this, then this is like finding the best book in the library of congress.
When presented with this problem — Users do not strive to guarantee success but to maximize the likelihood of satisfaction.
- Discovery: As we discussed, the problem of plenty could lead to drop-offs. There is another problem though, when we turn to a product, typically we have a problem — this could be a problem of feeling bored, not having the right tool etc. And when we are stuck with a problem, we try to turn to someone — a friend, a guide, an expert who could simplify our lives for us. This is why when we need an answer we turn to Quora, when we need quick information about something we use Wikipedia, and when we have to watch movies we turn to Netflix. The last thing we would want from a friend, a guide, or a product, in this case, is to spew 1000 results and leave us to figure out what we want. This is why most products are moving away from searching experience towards more of discovery experience. Case in point — Instagram, Pinterest, Netflix, Apple News etc.
So Discovery, the next stage in the funnel is the product’s opportunity to be the guide, the authority, and bring the few inventories that the user would love. This is done by a blend of reading various user attributes, understanding likeness scores of inventory, likeness & differentials of various users, and some data science magic — I will write an article on this separately.
- Evaluation: No matter how many choices the product presents to the users unless the product just returns one — which has its own issue, users will still need to evaluate. The costlier the item for example purchasing a house vs watching a movie — the more the evaluation time and the more are the people involved in the decision systems. In a behavioral study, this is also called persona or user role split. Providing an easy means to compare various attributes of the choices is what makes the evaluation stage interesting. However, very few products focus at this stage of the funnel. This could be because of letting users take a call — but it would not hurt to simplify the decision process — would it? At worst, users will ignore the cue, at best users will move to the next stage in the funnel, which by the way is the most important stage of the funnel.
- Booking: This is the last stage in the funnel, also the most important. I tagged this as Booking but this could be purchase, subscribe etc. depending upon which product we are talking about.
- The above were the stages in the funnel as you would like users to traverse the flow. However, the user journey is rarely linear, and most users keep falling off the funnel. This is why we need some way to bring users back into the funnel through various means such as Email, Facebook re-targeting, Push notifications etc. These channels act as a reminder to the user to keep coming back to the funnel. When users come back to the funnel — they are called returning users. If they have already purchased the product and come back then they will be called the ‘repeat users’. Building an efficient conversion funnel and in turn, an efficient growth engine needs channel strategy to bring users back to the funnel with relevant messaging and appropriate timing.
The objective of the conversion (CRO) engine is to come up with a game plan to capture the intent — Internal motivations of the users.
As noted above, the objective of the CRO is not to generate intent in the users. You would expect users to already have some intent if the acquisition engine has done its job right. The objective of CRO engine, thus, is to create an experience where the product is invisible to the users and help users move from one stage in the funnel to the next by capitalizing on the internal intent. This is why CRO requires understanding product frictions — many a time added by the core product team to achieve a certain objective, through user research and behavioral studies on the platform and remove them to create a seamless transition through the funnel.
In the next article, we will discuss the other two but really important engines — the pricing and the incentive engine. Without talking about those, we can't complete the growth engine — not a sustainable one at least. Read part II here.