January 15, 2023
20 mins

How to build successful marketplaces & platforms

Originally published in April 2018 here

Building a successful marketplace has some peculiarities compared to building a “normal” product or service-based business/venture. Contrary to popular belief the hardest part is not necessarily to solve the “chicken & egg” problem. The hardest part is to find a “marketplace model” which pleases all participating parties: buyers, suppliers and the platform operator (=you) in the long term. Most marketplaces do not fail because the founders are unable to solve the chicken & egg problem but rather because the chosen marketplace model itself is not a good fit to the realities of the targeted market. Not being able to solve the “chicken & egg” problem is then just a symptom of a unfitting marketplace model and not the root cause. The goal of this article is to help you think through your marketplace model based on the realities on the ground of your targeted market.

On a high level, marketplaces need to succeed in the same 4 dimensions as any other successful venture. My current employer BCGDV uses the following dimensions to evaluate successful businesses in the context of building ventures: Desirability, Feasibility, Viability & Strategic Value. So what does this mean in terms of building marketplaces?


  • What is the value proposition for the buyers? (buyer meaning the party spending money)
  • What is the value proposition for the sellers? (seller meaning the party receiving money, e.g. supplier)
  • Is the targeted market itself a good fit to build a marketplace at all and therefore desirable for the marketplace operator?


  • Which approach is best suited to solve the chicken & egg problem?
  • Based on the chosen approach, how to do and scale marketing?
  • How to support the transactions between buyers & suppliers in a way that both sides will continue to come back to the marketplace?


  • What is the typical ticket size of a transaction on your marketplace
  • How can you best capture a part of the value you bring to buyers & sellers?
  • Is the targeted market big enough (& ideally growing)?


  • Besides making money, does owning/operating the marketplace give any additional strategic advantages?
  • Can you use any unfair advantage from your (corporate assets) to make the marketplace successful? Does the marketplace give you an unfair advantage in other areas of your business?
  • Do the potential buyers & sellers have a strategic reasons not (!) to join your marketplace?

The term “marketplace model” is the combination of the answers of all of the above. You only have a good marketplace model when you check all the boxes above. Let’s go through them one by one.


As discussed the marketplace needs to be desirable for all participating parties: the buyers, the sellers and the marketplace operator.

A buyer typically looks for the following in a marketplace:

  • A faster way to find sellers (e.g. due to curated/ranked lists)
  • More trust in the sellers (e.g. due to ratings)
  • Price transparency
  • Minimum service levels
  • Standardized comparison options
  • Access to larger sellers which I could not contact directly (e.g. Alibaba)
  • ...

Sellers typically look for this:

  • Easier / cheaper to acquire customers
  • Access to new types of customers (e.g. Alibaba)
  • Make more money (e.g. through better prices)
  • Less operating costs (compared to running a shop themselves)
  • More information about the buyers (e.g. through analytics)
  • ...

The list above are just examples, for your specific marketplace you need to conduct in-depth interviews with buyers & sellers to find out what they are looking for and what their red flags are.

The main point I want to make here is that buyers & sellers normally don’t want the same thing in marketplaces. Some of the things buyers & sellers want are even mutually exclusive, e.g. sellers don’t want price transparency - in general they don’t want to make it easy for buyers to compare prices because this will lead to decreasing prices in their markets. On the other hand buyers want exactly that.

Therefore a good marketplace always needs to consider the realities of the targeted market. You need to differentiate between a “buyer’s market” and a “seller’s market”. In a buyer’s market the buyer can rather choose which seller to take (e.g. a buyer in a supermarket). In a seller’s market the seller can rather choose the buyer (e.g. FIFA World Cup tickets). So the different players have different power. You need to build your marketplace model in a way that the side gets more of what they want which has more “power” in the market currently.

Both sellers & buyers will look at the overall package: for example if you have a lot of buyers on your platform, then you can make the platform really bad for sellers and they will still come. We will discuss in the feasibility section which options you have to finetune this and make it work.

As a rule of thumb: don’t build the marketplace to change the power dynamics in the targeted market, rather build the marketplace to reflect the power dynamics of the current situation. If you want to change the power dynamics consider introducing new players into the market (e.g. UBER did not use the existing taxi drivers (who have too much market power) but rather used new drivers/sellers).

For the remaining part of this section I want to focus on the question how a market needs to look like to make the marketplace operator happy. You have a higher chance of building a successful marketplace if the market has the following attributes:  

  • The marketplace increases the pie of the target market instead of just replacing existing transactions (e.g. Foodora increased the market for home delivery, other examples include Etsy, oDesk, ...).
  • Solving a real pain point on at least one side of the market (e.g. really hard for buyers to find a potential seller).
  • Stable “power dynamics” between sellers & buyers so you can finetune the marketplace model accordingly. Shifting power dynamics make this harder (e.g. in seasonal businesses)
  • Marketplace becomes better when more transactions have happened (e.g. through ratings).
  • Network effect, the marketplace/platform becomes more powerful the more people participate (e.g. facebook which gets more valuable if more friends participate).
  • The content / items listed on the marketplace have a long “lifetime”, e.g. a housing listing “lasts” a few months vs. a demand for an UBER cab which needs to be fulfilled within minutes
  • High usage frequency, so using using the marketplace becomes a habit and people recognize your brand
  • High ticket value
  • Big & growing market
  • Cheaper customer acquisition costs than acquiring the customer directly (e.g. Check24)
  • Integrate deeply into existing processes and/or habits to make switching away hard
  • Be able to bring value on top of the pure matchmaking of buyers & sellers. So people don’t bypass the marketplace once they have found a seller (e.g. “Book a Tiger” problem where a matched cleaning persons will be hired directly after a few months).
  • Products/services from the sellers should be highly standardized and usable from all buyers but otherwise really hard to find (e.g. content on YouTube, UBER riders). If you have a very differentiated & customized product/service a marketplace is normally not an attractive offering.
  • Number of sellers for a comparable product should be high - if you have 100 sellers fighting for your buyers you can get a higher share of the created value than if you only have 1 or 2 sellers.

The more of those characteristics your target market has, the higher will be the lock-in effect of your marketplace. You can still make it work without some of these characteristics but it will be harder. If you don’t tick any of those boxes above you will have to actively address them in the definition of your marketplace model. I would recommend to treat the list above as a checklist with things to consider when you further define the marketplace model in the following sections.


So now we have an understanding what buyers, sellers & marketplace operators are looking for. The next question is how to make it work and bring it all together. So which approach do we choose to bring enough sellers & buyers on the platform, how do we scale this approach and how do we actually support the transactions of those buyers & sellers?

So far I have seen the following approaches to bring enough sellers & buyers on a platform.

Es wurde kein Alt-Text für dieses Bild angegeben.

There is no “best” approach - as discussed it depends on the realities of the market which one to choose. You should also use your strategic “unfair advantage” if you have any. For example corporate partners and/or venture capital investors sometimes have market access, brands or tech available which they can & should leverage to make the market work.

Once you have decided on your approach, you can start doing marketing based on that approach. Depending on the CLV of your buyers & sellers you will choose any mix between those outbound marketing options to attract buyers & sellers:

Es wurde kein Alt-Text für dieses Bild angegeben.

Your outbound approach should very likely be complemented with inbound marketing, e.g. by creating content which can be easily found on the web (SEO strategy).

I won’t go into detail in this article how to do different types marketing / sales and how to scale them properly - this is a topic for another article. For the moment here are only some general best practices to consider when scaling a marketplace with marketing:

  • All strategies are easier when you chose a focussed approach, e.g. start at a single university (like facebook did with Harvard) or a specific industry, ...
  • Piggyback if possible - meaning use somebody else’s platform and then extend from there, e.g. what Paypal did with eBay
  • See if you can somehow “fake” supply on the platform by also listing sellers which are not yet on the platform officially. You sometimes can still list their names & publicly available contact addresses
  • Every new channel will always be expensive at the beginning, then you optimize and then it gets expensive again once you start scaling
  • Think about how you can minimize adoption hurdles, e.g. don't try to change the user's behaviour or at least as little as possible. Rather piggyback on existing behaviour.
  • Think about how you can you reduce switching costs
  • ...

Let’s now assume we have selected the right approach and were able to bring enough buyers & sellers on the platform with our marketing approach - how should we actually support their transactions now?

The main things to define here are:

  • How can buyers indicate what they want
  • How can sellers indicate what they sell
  • How exactly your marketplace supports the transactions between those 2 players after they found each other

Options for buyers to indicate what they want are in general some type of search (e.g. Google, Matmatch.com, …) or creating some kind of classifieds / offerings (e.g. on dating apps). It’s not always the case that the supply creates an offer and demand searches for it. Sometimes (e.g. MyHammer.de) the buyers enter their wishes as “classifieds” and then the supply tries to get the deal.

You will choose one or the other option depending the type of market. If its a buyer’s market you want to make it as easy & straightforward for the buyer. UBER for example shines here: as a buyer you just have to press a button. Other good examples include “Same day delivery” from Amazon or the “buyers protection” from eBay. Getting this experience right is a critical success factor for buyer’s markets. Alternatively if it’s a sellers market you want to make it as easy as possible for the seller to put their offers in the market, e.g. OpenTable.com: since the good restaurants don’t really have a problem to get their tables full, they will only use the service if it also provides standalone usage for them. They then get the marketplace without additional operational effort.

So again - how exactly you support the transaction between buyer & seller depends heavily on the type of market you are working for. Here is a list of the different types of transaction support I came across in marketplaces:

Es wurde kein Alt-Text für dieses Bild angegeben.

The choice how to support the transactions is where the magic happens in marketplaces. What information do you show the buyers & sellers? How much do you integrate into the transactional process? How much pricing do you show? Does the payment happen on the platform or outside? (rule of thumb: Buyers market: Payment should happen on platform, Supplier’s market: payment will happen outside of your marketplace).

You also need to make sure you fit with the mental model & processes of the current market. The people on the ground need to use it. Not only the company in general. You need to find a reason for the actual person using the tool to use it ideally every single day. It is not enough that the company (the employer) would be making money if they would be using it.

A full feasibility check also includes a technical & legal perspective on the problem. So what can you build as an MVP in the first 3-12 months. What are legal requirements? However since I want to focus on the peculiarities of the marketplace I will again have to skip this for the purpose of this article.


So now that we have a marketplace which buyers & sellers want and is feasible to build: how can we make money out of it as a marketplace operator? The good news is: in general you will always find a way to monetise the platform as long as you manage to bring buyers & suppliers together. It just depends on how much of the value you can capture with the choice of your business model.

On the revenue side the business model should follow the selected type of transaction support (see feasibility). If you have payment on your platform you should probably get a share of the transaction. If you only do lead generation you will get some kind of lead fee or advertisement revenue. The higher the lock-in of the marketplace the more value you can capture. The lock-in effect is bigger the more your marketplace checks the boxes described in the desirability section.

If you don't have an easy payment process integrated yet it is very likely that a subscription model is better for you. Even though the "sale" is easier when your offer is free (or only transaction based), experience shows that making your customers pay for your platform makes them pay more attention to your platform and it helps them to take it more seriously (German mindset: "Was nix kostet, ist auch nix"). Basically once the boss was convinced that the company will pay for the service he is also more likely to encourage his/her employees to use it. A subscription based model makes the initial sale harder but those customers are then real customers. Ideally have your customers sign an auto-renew 6 or 12 month contract with a free trial which they can cancel in the first month.

You also need to look at the “basket” size of a typical transaction and how many transactions are happening overall as well. Consider that the marketplace may seriously impact the market itself over time. A few years ago the market of food delivery from restaurants was almost zero. Now due to Foodora, Delivery Hero and other players it is hundred of millions EUR in Germany.

On the customer acquisition cost side you need to make sure the marketing approach chosen in the feasibility section scales within reasonable costs. This very likely means that you will have to systematically test out different channels to get buyers & sellers (see feasibility for marketing mix). Don’t just look at the cost to acquire those users, you also need to consider the quality of those users. Inactive users are worthless. You want active participants in your marketplace which generate transactions.

A good test if you have a viable marketplace model is that you can tell me your main KPIs you will use to run your business next to revenue & costs. What are the KPIs you check every day? What KPIs do you put on the wall in your office? What are the main goals of your product team, your sales / go-to-market team? What means success for your business? Selecting the right KPIs is again a big topic which I won’t be able to handle in this article - however I can recommend to read this book here: Lean Analytics. For now again just a few best practices:

  • Don’t use vanity metrics like number of pageviews: They will always go up and are not actionable. Rather focus on engagement, e.g. what’s the share of people coming back every day? Retention (do they come back?) & engagement (how often do they use the service?) is key to most digital businesses.
  • Don’t confuse correlation & causation: always try to find the root case. Use ethnographic research to understand reasoning behind certain behaviour. Only then you can understand why KPIs correlate
  • Always look at cohorts: The quality of your users on your marketplace will change drastically over time depending on your marketing activities or just purely randomly. Don’t freak out if your KPIs change over a limited time. Keep calm & carry on. Wait for statistical significance before you take action.
  • Marketplace are a funnel optimization problem at the end of the day. Define certain steps across your funnel (e.g. buyer visited, searched, found a seller, transaction happened) and then relentlessly optimize this funnel. Start at the beginning of the funnel. Do this for both buyers & sellers. This will then also inform the CAC & CLV of your customers together with your retention numbers which will decide your viability.
  • Start with your current understanding of your success KPIs. You can have different KPIs for different parts of your company which complement each other (e.g. CEO Level for overall KPIs, CPO more product focussed, COO/CSO more BD focussed). Change them if you learn more. Having less KPIs to look at (1-3) is better than having more (7-10). Having one main KPI which everybody looks at is great.

I will leave you with the eternal truth of any digital business: you have a viable business if your customer acquisition costs (CAC) + operating costs is smaller than the customer lifetime value (CLV) of a typical customer. It is as simple as that :)


The strategic dimension is especially important if you are a corporate partner or VC with a bigger view on the market. If you are a (corporate) investor you should consider in the feasibility phase if you can use any (corporate) assets to make the marketplace successful.

If you are a normal startup the strategic dimension probably seems less important. Just start, get traction and go from there. Still you might want to consider these questions to understand how other players on the market might see your marketplace:

  • Do the potential buyers & sellers have strategic reasons not to join your marketplace?
  • Besides making money, does owning/operating the marketplace give any additional benefits? e.g. giving you an important market position (which is not taken by a competitor) or making important data accessible for one of your other businesses?
  • Does the marketplace give you an unfair advantage in other areas of your business?


At the end of the day building a platform/marketplace is easy - you just need to make all members on the platform happy: buyers, suppliers and yourself as the marketplace operator ;)

Finding a good marketplace model requires decisions on several dimensions in desirability, feasibility, viability and needs to consider strategic implications. I hope this article helps to give a structure to think through your marketplace based on the realities on the ground. You can use this document like a checklist following these simple steps:

  1. First look at Desirability to understand the market and its characteristics. Is is a buyer or sellers market?
  2. Based on that look at Feasibility: Choose the approach fitting to the market (and use your strategic unfair advantage if you have one). Decide how to scale and how to support the transaction between buyers & sellers in detail.
  3. Based on that chose your business model and check its viability.
  4. Consider the strategic dimension last. While you should use your unfair advantage to make the marketplace work (see feasibility), you very likely don’t have a viable marketplace based venture if you only see strategic reasons to operate it.

You will only be successful if you manage to be successful on all these 4 dimensions. If on the way you learn for example that viability does not work out, then you need to go back to feasibility and change your model. It is an iterative process.

As a last piece of advice: you should start by making your buyers & suppliers happy. Only after you have traction you can make yourself as the marketplace operator happy. You can evolve your marketplace once you have solved the hen & egg problem at the beginning and started gaining market power. Then you can start dictating the rules and enjoy the good life :)


Sometimes buyers & sellers won’t join your marketplace due to exactly this strategic reason. They are afraid that you will change the market once your marketplace has gathered market power. At the end of the day this is what almost all successful marketplaces have done. However there may be a potential solution by using the Blockchain which could give all parties on the marketplace a say in how the marketplace will evolve. But this is again a topic for another article.

PPS: here is a great overview about other good marketplace articles. This article here is also pretty good: https://medium.com/@sarahtavel/the-hierarchy-of-marketplaces-introduction-and-level-1-983995aa218e