Scaling a Marketplace Product

Scaling a Marketplace Product
Paul Zeckser led product ad HomeAdvisor from 2011-2020 and is now the Founder of Fresh Product Design, a full-service product strategy and development consultancy. In this guide he lays out how product, sales and marketing and operations should coordinate to scale marketplaces. He covers metrics to track, how to think about matching, and guidelines for monetizing marketplace transactions.

This is a preview of a single guide - to access the full library or request a call

Questions covered in this guide

What are the different types of online marketplaces?

B2B vs. B2C vs. B2B2C – who is the user? You’re going to approach each of those marketplaces pretty differently depending on the user types and the frequency with which they’re interacting with your product. 

Commodities vs. differentiated services – e.g. a once-in-a-lifetime kitchen remodel vs. booking a dinner reservation. For a kitchen remodel, it’s extremely important to you as the user to see what’s the skills and abilities of the different options you could choose in that marketplace. If you’re just in a marketplace to get from point A to point B or to book a reservation in a restaurant, then the platform is much more of a commodity.

Also consider:
  • Frequency of use – overall, and on supplier and buyer sides of the marketplace. For example, an Uber driver might be using the app vs. a casual consumer might only be using it once a month.
  • Price point – you’ll design your marketplace differently if expensive goods or services are being sold, and there’s a high expectation of quality. 

Who are the different product and ops team members who work together to manage a marketplace? What are the options for structuring the org?

Who’s involved:
  • Product – responsible for the product roadmap, can be the overall lead
  • Operations – responsible for human touchpoints around the product
  • Sales – usually primarily responsible for recruiting suppliers (or potentially large buyers)
  • Marketing – usually primarily responsible for generating demand on the buy-side
  • Finance – responsible for tracking revenue and costs
  • Legal – responsible for compliance

One org option: product drives strategy – product takes the lead on what’s going to be built, why it’s being built, and the order things are being worked on. Product is the central receiving point for inputs from stakeholders within the company, and on both sides of the marketplace. In this case, product is responsible for coordinating with operations, sales, marketing, finance, legal, etc. The goal is a coherent experience for the end-user. 

Alternate org option: business operations sits alongside product – product and business operations (or a general manager) form a kind of marriage, and like a marriage who owns what will depend upon the strengths of each leader or team. Often product owns the product roadmap, technology and data, while the general manager owns revenue objectives, sales and operations.

What are the key responsibilities of a marketplace product manager (especially as they differ from other software PMs)?

The key difference is that you’re thinking about more than one customer group – for example, if I’m the product manager at Airbnb, and I don’t have both a great experience for the property owner as well as the renter, it’s going to fall apart. 

You need to manage when they’re in conflict – a lot of times, interests between both sides of users will align, but when they don’t, there’s a classic debate of “are we a marketplace that’s going to prioritize the needs of one user over another or are we going to attempt to balance the needs of both?” It’s a specific challenge that’s unique to software marketplace PMs.

What key metrics should you track within a marketplace?

You can’t run marketplaces without metrics – no matter the size of the marketplace. For anyone who wants to build out a large marketplace, the product decisions are going to have to ultimately be driven almost entirely off analytics, with a small role for qualitative feedback. 

Track acquisition metrics on the demand side:
  • Cost per buyer acquisition 
  • Value per buyer acquired
  • Number of buyers acquired
  • Types of buyers acquired (by segment, by geography, etc.)
Track availability metrics on the supply-side: 
  • Enough suppliers – who’s going to fulfill your services? Do you have enough of the supplier types you need?
  • Proper inventory – do your suppliers have the proper inventory? E.g. for a marketplace like Etsy, there’s a limited inventory per supplier.
Track match quality metrics – there’s a certain percentage of time where you bring two parties together and they align on all the dimensions they want to match with each other on. Then, there will be other times when you bring someone in and what they’re looking for can’t quite be fulfilled by the supply that you have available. You want to use metrics to understand the quality of the matches. 
  • Match frequency – how often do you successfully make matches (that lead to transactions)? These lead to promoters of your brand.
  • Strike-out frequency – how often are you unable to make a match? You want to measure these instances because there are usually the ones that lead to brand detractors.
Track service or product experience metrics – after the connection is made, does the experience continue to be good?
  • Timeliness – is the service provider always on time or did they get back to the customer quickly? 
  • Customer satisfaction (e.g. NPS) – is the buyer happy with the quality of the service? For example, how many times does the lawn care service leave and the grass looked great, and how many times did they miss a spot?
  • Real-world impact – find a way to track when something goes wrong that you’re not able to see from the digital side of things, but that has a real-world impact on the users of your marketplace.

How should you choose and use a North Star metric?

What are you optimizing for? – are you optimizing for demand growth, supply growth, or the quality of the connection and the experience? From there, you can develop your Northstar metric. Ultimately, I think you’re optimizing for that high NPS or customer satisfaction score because it drives repeat usage, and that’s a powerful thing in marketplaces. 

It’s hard to get far without good matches – so you need to always be optimizing for match quality. As you start to do that, you’ll often find that it’s easier to bring in one side of the marketplace over to the other.

Optimize for the side that’s harder to bring in (and that side might change over time) – e.g. for the home advisors it can depend on the economy. So, in a situation where the homeowner market is really bad and homeowners didn’t feel like they could spend a lot of money, then a lot of professionals would be coming to home advisors because they can’t find any business. In the opposite situation, there’s a lot of demand from consumers and they’re going to home advisors because they need help finding professionals who’re free. 

Over time, you might need to focus on different things to optimize the match value – you’re always trying to optimize for the quality of the connection, but over time, you might need to do something different to achieve a great connection.

What type of data analytics should marketplace companies set up as they scale?

You need a product analytics or data science team that’s involved across teams:
    To inform sales – and help drive outbound phone calls to specific geographies or specific regions. 
  • To inform finance – on the value of each acquired user, the cost of acquisition, etc. This helps you know how much you’re making in different markets in different categories. 
  • To monitor the health of the ecosystem – for example, Uber drivers can turn off their app at any time, so we’d want to understand when drivers are more or less likely to turn off the app and how that aligns with when the most people are looking for a ride.

Data scientists build models that feed into your matching algorithm – engineering often co-owns the matching model with your data science team and to drive satisfaction on sides of the marketplace, and match quality. Use data science to find ways to generate more matches. For example, at Home Advisor, we’d attempt to be predictive about storms, such as hurricanes on the East Coast. We look at past hurricanes and what they had done to our consumer demand and supply. Then, when we saw that another storm was coming, we’d start proactively notifying contractors in adjacent areas and adjacent states that they might want to increase their supply or their profile coverage in those areas. 

When you first introduce a data role, look for a jack-of-all-trades who complements your team – they’re highly analytical, they’re comfortable with writing algorithms, and they like constantly being immersed in data and building predictive models. 

As you scale, you might have one data scientist for every 1-3 delivery teams – bring the first one in at 10-15 people in product and engineering, then you may not need the second one until you’re at 25-40 people.

What are the different options for monetizing a marketplace? How should you think about setting and adjusting prices or commissions?

The buyer is on a journey through your product, the further they get, the more you can monetize – because the further they go, the more value you’re matching them to. If you decide to monetize earlier in the experience, you monetize at a lower rate and if you decide to monetize later in the experience, you’re monetizing at a higher rate. 

Lead feeds are at the low-touch, low-value end of the spectrum – if you’re just providing leads, there’s a low probability that any one lead will generate revenue for a supplier, and you can’t charge a lot for them. 

Connecting two parties is more valuable –  if you connect the parties through a conversation or an appointment scheduled online, your close rate is going to be much higher and you can monetize at a much higher rate. 

A win-fee model is the most valuable – if you’re only charging the supply side of your network when they’re making money, you’re proving the most value, and you can charge the most. E.g. Uber only serves up already-committed riders to their drivers, and can therefore take a high commission.

How should marketplaces think about facilitating and monetizing payments between buyers and suppliers?

Buyers often like the marketplace to take payments for security – if it comes with certainty or reassurance. There’s value in the marketplace taking on responsibility, e.g. saying “if the provider cancels on you, we’re going to find someone else”. 

Suppliers like marketplace payments for easier collections – with home services, for example, some providers will leave when the job is done and not try to get payment from a client for 15 days or so and collection becomes a nightmare. So, with the platform, they can just say that they did a job and then collect their payment. 

The more potential distrust between buyer and seller, the more sense payments make – if either or both parties are worried that the other won’t perform a high-quality service, ship the product, pay on time, etc., the marketplace can reassure them by putting some guarantees in place. In exchange for taking on that risk as the platform, you can justify taking a percentage of the transaction.

Marketplaces can monetize payments in multiple ways:
  • Taking a cut in exchange for “insurance” – the involved parties will be alright with the platform taking a certain percentage cut since the platform is taking on a certain level of risk. Make sure both parties understand the value you’re providing to them too so they don’t question the cut you’re taking. 
  • Offering payment terms – if suppliers want their money faster, they can pay a little more to get it (e.g. a higher commission to receive funds in 5 days vs. 30 days).

Which teams are responsible for driving the buy-side of the marketplace? What tactics are common for marketing to and driving usage among buyers?

Teams involved:
  • Marketing – responsible for bringing buyers to the marketplace, and uses all kinds of tools to bring people in, both with paid and organic leads. 
  • Product – responsible for converting and bridging potential customers into what the product experience is going to be. 
  • Specialty optimization role – this role sits somewhere between product and marketing. They’re responsible for squeezing every drop of conversion out of the funnel. They also have the flexibility to work outside the standard product release cycle. 
  • Operations – supporting potential customers through live chat to make sure they don’t get stuck on the website. Ops might also support existing customers who’re in their account and they’re struggling to figure something out, and should be focused on cross-selling and upselling existing customers.

Marketing and product need to collaborate to create a seamless funnel – the messaging between marketing and product needs to be the same because you don’t want marketing to bring someone to the website with one set of messaging, and then they have a jarring experience once they enter the actual product.

Which teams are responsible for driving the supply side of the marketplace? What tactics are common for recruiting and suppliers?

Teams involved:
  • Sales (varying from self-serve to consultative) – some sales can be almost entirely self-service, like Uber. People download the app and then upload their information. On the other end of the spectrum, you might have a sought-after provider that you need to recruit onto your platform. For example, a successful remodeling professional that does $5-10 million a year in business and you want them to join your home services platform. There’s no way they’re going to get on the platform in a ten-minute signup process online; they’re going to need to have multiple conversations with a highly skilled sales professional. 
  • Supply-side marketing – marketing plays a huge part in driving warm sales leads; you might have a dedicated team focused exclusively on supply. 
  • Product – building self-serve flows or internal digital tools to support the human experience. If you’re creating more of a self-serve signup experience, then that’s a huge product and engineering effort. 

You need to make it really easy to work with you if you’re only a fraction of your suppliers’ business – you need to make it easy for people to sign-up and use the product by doing things like integrating your platform with other tools your suppliers use. If you make it difficult for them, then they’re not going to want to work with you.

How can you prevent disintermediation?

Teams involved:
  • Guarantee payments – can you guarantee that someone will get paid in a certain amount of time? 
  • Provide downside protection – can you guarantee the consumer that they’ll receive a certain level of quality or experience? If something goes wrong while the consumer is using the platform, you have to show that you’ll be there for them in a moment to fix it for them.

What steps can you take to ensure consistently high supplier quality?

Establish vetting standards – you need strict minimum criteria. Even if someone has a lot of inventory they could bring to the marketplace, if they don’t meet the criteria, then they aren’t worth having as a part of your platform. 

Reward good behavior and penalize bad behavior – identify the few behaviors that create a great experience (e.g. timely responses), monitor those, and reward them. The behaviors can be rules that you write manually or that you create with an algorithm. At the same time, de-prioritize suppliers who exhibit behaviors that lead to poor buyer experiences or worse match rates. 

Create some intervention for the group that’s struggling – particularly if you’re supply-constrained, find ways to help struggling suppliers improve. Provide them with tips or tools that you’ve created to help them find success in the marketplace.

How can you continue to grow the top line when faced with supply limitations?

Support the creation of more supply – if you’re in a particular industry, you can help support or fund or provide scholarships for training in that space. For example, we often talk about the number of unfilled trade jobs out there, and there are a lot of trade schools, so you could help people attend those schools. When they graduate, maybe they get a free account on your platform and a $500 credit.

Become a player in your own marketplace – you can establish your own supply of key services or items through acquisition, or by organically building your own teams. 

Offer premium pricing, shared between vendor and marketplace – when demand is high, increase prices and share some of the surplus to incentivize increased supply, e.g how Uber has surge pricing when there’s a lot of demand for a ride, but not enough drivers.

What legal and compliance roles and tasks are important to establish? 

Seek legal counsel from the very beginning, and inside counsel once you can – marketplace law is a fluid space, so you don’t want to take on any risk without any kind of counsel. There can be a huge disconnect between what your terms and conditions are vs. what the expectations are. 

The bigger your marketplace, the more attention it attracts – if your marketplace is smaller, it’s a lot less likely that someone is going to come in and nitpick your legal decisions. 

As you grow, have an audit conducted by outside counsel and factor it into your roadmap – the lawyers will tell you what items they think you could get into trouble, and which ones aren’t as high of a priority. Work addressing the riskier items into your backlog so they’re being taken care of over time.

What are the most important pieces to get right?

Create a culture of ownership in product and engineering – the people building the technical product should feel ownership over the quality of the experience, the relationship with end-users, and creating an experience that delights those users. Empower engineering by saying “here’s the big vision we want to get to, and your team needs to figure out how to get us there.” 

Focus on the “magic” matching experience – get the entire org focused on how you create the seemingly magic match that makes a user a promoter of your brand and makes them want to tell everybody about their experience.

What are the common pitfalls?

Don’t assume you’re a typical user – never stop talking to your users or stop looking at metrics and paying attention to what people are doing on your website. You might start to think, “well, I’ve been around the business for a while and I use this marketplace with a fair amount of frequency, so I have a pretty good view into the truth.” You’re probably not the typical person who’s using your marketplace, so you need to keep an eye on what your users actually need and use the data you collect to prioritize your backlog. 

Don’t get distracted from user experience by internal asks – don’t change the goals and priorities of the roadmap because legal asks you do to something, or finance asks you to do something, or the board asks you to do something, or because you’re scared of a competitor. Those asks can send you down a path where your roadmap is totally derailed, and you’ll look back at the end of six months and realize that you didn’t improve the user experience at all.

To browse the full library of guides

Already a member? Sign in to view the full library.


Scroll to Top

Request Access


Want free guides?

We feature guides every week in our newsletter