Parking One Side of a Marketplace

Building a marketplace is hard. It’s difficult because you’ve got two different customers: buyers and sellers who are typically two different people (exceptions include classified ads and dating apps).  Two different customers implies:

  • Two different value propositions
  • Two different routes to market
  • Two different products
  • Two different support structures
  • Two different EVERYTHING!

Marketplaces are two businesses running simultaneously and both have to succeed for the overall operation to prosper.  This results in marketplace failure rates that are much worse than a normal startup, however they have the benefit of being able to scale wildly if they are successful.

I’ve dealt with a lot of marketplaces over the last year. Everything from a private tutoring market in the Middle East, rental accommodation in Ireland to dog boarding in Europe. After ten years in WhatClinic, people think I’m some sort of expert. I don’t have any secret sauce, however I do know a lot of the pitfalls.

The first pitfall is trying to build both sides of the marketplace before you have sufficient resources / team. Even in a successful marketplace with teams supporting both sides of the business, it’s hard for the CEO to wear two hats; it’s nearly impossible when resources are very limited. Focus your resources on one side of the market first.

Recognise who your primary customer is. In general, this is the side that has fewer problems finding the other party . This is the side of the market that you have to crack. It’s where the real risk is in the business – if you can crack this side of the market, the other side is simple in comparison.  

So for WhatClinic the primary customer was the consumer not the clinic, for Google Ad Network it’s the advertiser not the partner website, for Amazon it’s the consumer not the seller. Uber, HostelWord, Hotels.com & Opentable all built supply first, parking the consumer side until they were ready [Note some edits made here from original posting thanks to comments from Ronan Percival)

  • If Amazon has has loads of consumers it is safe to assume they’ll be able to sign up vendors
  • If Google Ad network has loads of advertisers then partner websites will flock to them.
  • Uber reckoned that if they sort out supply that demand would exist

But it’s a chicken/egg (horse/cart) problem. It’s difficult to build one side without the other because no value is being delivered. What’s the answer? Try and park away the second side of the marketplace by directly addressing the primary side’s value proposition without building a market on the other side. 

Can you park away one side of the market?

  1. Offer the service directly yourself. So if you are an online babysitting marketplace you can directly hire a few babysitters in your first city. Once you have proven the value proposition for the parent and have more demand than supply, you can switch your attention to building the other side of the market, targeting the sitters.  Amazon was an online store before it became a marketplace.
  2. Seed the market. Get one prestigious name on the secondary market that will drive large volumes from the primary market. For example if you are a jobs board then a job for Ferrari or Dolce & Gabbana can drive tens of thousands of CVs.
  3. Build the supply without the relationships. For example if you are a real estate marketplace you can simply list properties without developing the paying relationships with the real estate agents. Once you have the primary audience you can start to build the relationships
  4. Build the primary audience for a directly related informational need. If you are an investment real estate marketplace you could build your primary customer audience by publishing research information before building the market.

My final thought is try and recognise when the marketplace exists because of a customer need for a market or because of internal business requirements. Uber’s riders & drivers don’t want a market (the rider just want a ride and the driver a fare) but Uber does. They don’t want the hassle and legal liability with directly employing drivers. However the consumer clearly wants a marketplace for homes to buy as they want to compare lots of options.  Having clarity on exactly who the marketplace serves should give you insight on how to get it started.

Rule of Thumb for the Direct Sales Channels that are Viable for Different Revenue Levels

So these are very general rules of thumb: there are loads of exceptions and as always context is king. That said, I find these rules a useful shortcut when analysing possible direct sales models for companies. The rules are based on the annual average net revenue that your typical customer is responsible for:

  • Below €500 annually you can’t afford to call a customer and you must have a 100% automated sales channel
  • Below €5,000 annually you can’t physically meet a customer and you must either have an automated or inside sales channel
  • Below €50,000 you can’t get on a short haul flight to meet a customer
  • Below €500,000 and you can’t get on a long haul flight to meet a customer

Each of the different direct sales channels has radically different costs, and companies need to structure their sales operation so that they can expect to recover the cost of making a sale in well under a year.  

A typical inside sales rep will have an OTE between €40K and €70K and a fully loaded cost of roughly 150% of that. If they close 10 deals a month then the cost of making the sale is going to be somewhere between €550 and €950. Clearly if the resulting customers are only going to bring in €500 in net revenue it’s going to take a long time to cover the cost of making the sale.

A typical field sales rep will have an OTE of €60K to €100K with a fully loaded cost of double that due to the cost of travel. They should be closing about 7 deals a month, but they typically require a full time inside sales rep to source the leads and set up the appointments, taking the total cost of making the sale to somewhere between €2,300 and €3,500. So while it’s possible to run a field sales channel at under €5k net revenue, it’s difficult.

Once you have to get on a plane the whole dynamic and cost base changes dramatically. Now we are realistically starting to look at enterprise based selling. OTEs range from €70K to €150K (and up) with a fully loaded cost of about 250%. You can expect a maximum of a deal a month. Typically one meeting a week is reasonable – more can be done if you have particularly high customer density, but in that case you should probably be considering a local field sales force. So your total cost of making a sales ranges between €15K and €34K although it likely to be higher as a deal a month is on the high estimate.

Inter-continental travel makes everything even slower and even though OTEs aren’t much different, the number of meetings possible in a month shrinks to about 1 a month and resulting sales to maybe three a year. This results in an cost of sale ranging between €60K and €130K.

I find this a useful ready reckoner to quickly evaluate the available direct sales channels when analysing a company. It’s a shortcut, and I know that if something is anywhere close to the boundary it merits further in depth analysis as the peculiarities of a particular business or context may make a sales channel that initially seems implausible possible.  

A Great ROI is Not Enough

It’s easy to think that all you have to do is provide a big enough ROI to your customer. However this is not enough on its own – you also have to move the needle for the customer’s business. Your ROI might be fabulous but if you don’t move the needle for your customer then it’s going to worth your customer’s effort dealing with with you.

For example let’s say you are selling sales leads for second hand cars. You are selling the leads for €25 euros each and you expect 10% to convert to sales, so it costs about €250 for each car sold. The dealership makes about €750 per car, so the ROI of 300% is great. You’d naturally think that you could sell these leads to anyone, after all you are effectively selling €750 for €250.  But there is another dimension – volume.

Let’s say you only have 20 leads a month to sell. A small regional dealership that only sells 20 cars a month will happily do business with you because you are likely to boost their sales by 10%. However for a large chain dealership that sells 1,000 cars a month you are only going to grow sales by 2%. You don’t move the needle for them, and it’ll be hard to get them to do business with you because the potential upside isn’t interesting enough.

So how much do you need to move the needle by? I’m sure this varies a lot from industry to industry but in my experience it needs to be about 5%. That’s not necessarily 5% of the whole businesses revenue or cost – it’s the individual’s personal target. So if you are selling into a sales manager then it’s 5% of their new business, if it’s the facilities manager it’s 5% of the cost that they are trying to reduce, etc.

Below 5% you simply don’t move the needle enough to warrant their attention. However, your importance increases the more you move the needle.  I have found that at 20% you can even insist on fundamental changes to the customer business, such as changing their procedures or even business model. Simply put, at 20% you are so important that they can’t afford to lose you.

Monthly SAAS Contracts Suck

The appeal is obvious – anything that lowers the barrier for a potential customer to sign up has to be a good thing, right? It’s attractive when you are scrambling for sales, any sales, that month to month subscription without a long term contract has become the default business model for young B2B SAAS businesses.  Unfortunately it’s the wrong model for nearly all of them.  Personally I blame the Signal 37 guys at Basecamp for embedding this in start-up culture, but that’s another story.

Why is it such a Bad Model?

  • No upfront cash means as sales scale so does the level of investment required to fund the sales channel
  • Lack of customer validation creates an onboarding bottleneck and high early churn rates
  • The customer is not invested in your mutual success leading to more failed customers
  • Commission modelling and payment become messy leading to salesperson disatisfaction
  • Revenue modelling is more uncertain meaning that financing becomes more difficult especially debt

Caveat – Month to month contracts can work well if there is a purely automated sales channel.

Fund the Sales Channel with Sales or Investment

Imagine it costs you €1,200 to make a sale, i.e. it costs you €1,200 to find a potential customer and complete a sale. Now imagine that you only have €1,200 and your monthly subscription is €100. On Day 1 you spend your €1,200 and make your first sale. When do you make your second sale?

You make your second sale in Month 13, when you have collected the €1,200 to fund it from your first sale

Now think about the exactly the same scenario, except this time you charge a €1,200 annual subscription upfront. On Day 1 you spend your €1,200 to make your first sale. When do you make your second sale?

You make your second sale on Day 2 using the subscription collected on your Day 1 sale. By the time month 13 rolls along you’ve made about 230 sales – a 23,000% improvement over the first example.

I accept this is contrived, however the core message is valid – you either finance your sales channel with investment cash or your customers’ cash and it’s nearly always preferable to do it with your customers’. After all, earning money from customers should be your company’s core competence, not getting investment.

Lack of Customer Validation

If a salesperson closes a month to month subscription, then what exactly have they sold? Maybe they sold a three year revenue stream or maybe just one month’s fee that won’t even pay the sales person comp.  The only thing that’s definitely validated is that the customer is willing to pay for the first month. Should you even regard them as a customer at all? Maybe it would be best to think of them as being on a trial until they have paid a certain number of invoices and are actively using the product. This makes a difference because your customer success and onboarding teams will get frustrated and will rightfully demand that they not be handed half completed sales.

The Customer is not Invested in your Mutual Success

Okay, I’ll admit that getting a months subscription off of a customer isn’t nothing. It’s a lot better than a free trial. But what proof do you have that they are committed to your mutual success? The first month fee is a trivial amount and there is no long term commitment.  No wonder your onboarding team have difficulty getting them on the phone and engagement is weak.

Contrast this to the customer that signed up to a one, two or three year contract – there’s a committed customer. There’s a customer that will call you angry when things aren’t working. There’s a customer that will fight to make sure they receive value. There’s a customer that you want.  

Comp Models Become Messy

What are you going to comp a salesperson selling month to month subs on and what money are your going to use to pay them their comp? If you comp them just on the month’s sub then you’re incentivising them to find customers who are willing to pay a one month sub – not customers that will be with you for the long term.  If you comp them on revenue from the customer as it comes in then the reward is disconnected from the actions you want to incentivise. If you advance comp on the basis of future revenue you create a cash flow issue and de-motivational claw backs when the customer cancels.

Comping on an annual contract is easy, and paying the comp is easy if the customer pays upfront.

Revenue Models and Investment Become Difficult

Month to month contracts have no committed revenue stream by definition. Revenue models have to use average churn rates to calculate the following month’s revenue with ever increasing levels of uncertainty the further out you go. In contrast annual contract business only have to use average churn rates for revenue a year out. Having this contracted revenue stream makes financing, especially debt financing, easy when compared to month to month contract. This is all the more important as month to month contracting needs more financing to fund the sales channel.

So why do start-ups almost invariably opt for the month to month sub? Its because they are afraid to ask for the sale, so they attempt to minimise it. They don’t charge what they’re worth (more on this in a later blog post) and then they split the sale into the smallest segment possible in an attempt to make their pricing irrelevant to the customer – I’ve even seen companies split it into a weekly fee.

What if customers won’t sign up for an annual contract?

If you’re a B2B company with a manual sales process then there is a good chance that you just have to accept that you’ve got a longer sales cycle than you think you have. If you are sure that your potential customer will only sign up to month to month then accept that this is a trial period and after the trial period ends they move onto an annual plan. This division makes it very clear that during the trial period the sales process is still ongoing and you don’t count the client as a customer until they sign up to the annual plan. In this way your customer numbers stay clean and responsibility for the sales cycle is clearly defined.   

Usual disclaimer: Context is king, actual results may vary, etc.