Working Out Whether a Marketplace Channel Pays for You

Most articles about online channels eventually reach a claim about results. The only results worth anything to you are the ones your own numbers produce, so what follows is the calculation, the figures you need, and the two places it usually goes wrong. It takes about an hour with your own reports.

One piece of outside evidence is worth putting on the table first, because it frames why the question deserves the hour. McKinsey published research in April 2023 titled The Multiplier Effect: How B2B Winners Grow, drawing on its Global B2B Pulse survey of 3,862 sales and marketing leaders across thirteen countries. It found that 48 percent of companies gaining market share were selling through industry-specific marketplaces, against 13 percent of companies losing share. That is a correlation rather than a mechanism, and it covers B2B broadly rather than this industry specifically, but it is a large enough gap to be worth understanding rather than dismissing.

Start with what a customer is worth

Almost every mistake here comes from valuing a new customer as a transaction rather than as a relationship. Storage and destruction need separate sums, and running them together is the single most common error.

For storage, take average monthly storage revenue per box, add average annual service revenue per box divided by twelve, and you have monthly revenue per box. Then the number most operators have never calculated. How long does a box stay? Take boxes permanently withdrawn in the last twelve months, divide by average boxes in store, and invert for average life in years. Most operators are surprised how long it is. Multiply monthly revenue by average life in months, then by the average box count of a new account, and you have what a storage customer is worth over the relationship.

For destruction the sum is different and usually smaller. A one-time purge is worth the job. What makes it worth more is conversion, so the figure to calculate is what proportion of purge customers become scheduled service customers within twelve months, and what a scheduled account is worth annually. If you have never measured that conversion rate, it is the single most valuable number in this article, because it determines whether purge acquisition is an investment or a transaction.

Then work out what you currently pay for one

Add together, for the last twelve months: the fully loaded cost of everyone doing sales including salary, vehicle, expenses and a share of management time, marketing spend of every kind, and the value of operations time spent quoting and visiting sites. Divide by new accounts actually won.

That is your current cost per acquired account. It is almost always higher than expected, because sales effort is a fixed cost nobody attributes anywhere, so it feels free at the margin. It is not free. It is just not itemised.

Then price the alternative you are not considering

Most operators compare a marketplace against doing nothing. The fairer comparison includes the third option, which is building an online purchase path of your own.

Cost that out properly. Pricing logic that works without a salesperson present, a service catalog, scheduling tied to real route and warehouse capacity, payment handling, contract terms a customer can accept unassisted, and ongoing maintenance. Then add the marketing spend required to make anybody visit it, because a purchase path nobody finds generates nothing.

That total is the honest benchmark. A marketplace is not competing against zero. It is competing against a build, and the build is why almost nobody in this industry currently sells online.

Now the comparison

For each channel, work out cost per acquired account and divide by monthly revenue from that account. That gives payback in months.

For storage, payback decides it rather than acquisition cost. A channel costing twice as much per account but delivering customers who stay eight years is a better business than a cheap channel delivering customers who leave in eighteen months. Under twelve months is strong, twelve to twenty-four is workable where attrition is genuinely low, and beyond twenty-four you need real confidence in retention.

For destruction, apply a harder test. A purge should be profitable on the job, at your actual cost to serve including the drive. Treat conversion as upside rather than justification. If a purge only pays because you assume it becomes a scheduled account, you are relying on a conversion rate you have probably never measured.

Contract sharing changes one part of the sum

There is a mechanism in some marketplaces, Annex among them, with no equivalent in traditional channels, and it is worth understanding because it affects the input rather than the output.

When a customer specifies requirements such as capacity, geography, certification or particular protocols, the platform matches those against member capabilities. That means you can be considered for business where you had no prior relationship and would not have appeared on any list. In practice it is how an operator gets in front of a requirement that would otherwise have gone to whoever the buyer already knew.

The effect on the arithmetic is that it raises the ceiling on addressable volume rather than reducing cost per acquisition. Model it as access to demand you were previously excluded from, and be realistic that requirements you cannot actually meet are not opportunities.

Where this goes wrong

Two errors account for most bad decisions.

Counting inquiries as customers. Volume converts at a rate, and that rate is a property of your responsiveness as much as the platform’s. This is acute in destruction, where a buyer with a two-week deadline takes the first credible answer. An operator replying within the hour and one replying next afternoon will get materially different economics from identical inputs. Measure cost per acquired account, never cost per lead, and be honest about your own response times before blaming a conversion rate.

Ignoring route and overlap. Some proportion of what arrives would have found you anyway, which makes true acquisition cost higher than the invoice suggests. And on the destruction side, jobs outside your route can be unprofitable at any price. Segment by geography, discount the overlap, and exclude the work you should be declining.

What the answer actually depends on

Run this and the decision turns on two things, neither of which is the channel’s pricing.

Your attrition rate, because that sets lifetime value and therefore what you can afford to pay. And your spare capacity, meaning spare racking for storage and spare route capacity for destruction. A new account is worth substantially more when it fills space you are already heating, or a truck already making the journey, than when it forces a new one.

An operator with low attrition, empty racking and an under-used route can justify acquisition costs that would be reckless for one at capacity with boxes leaving steadily. Both are looking at the same platform and the same price list, and they should reach different conclusions.

Most operators, running these numbers for the first time, find they have more spare capacity and longer customer lives than they assumed. That is the common case rather than the exception, and it is why the arithmetic is worth an hour. The question was never whether marketplaces work. It is whether one works for your numbers, and now you can check.

Disclosure: Annex is affiliated with O’Neil Software. It is named here as one example of an online channel, and this article is not a recommendation to use it in preference to alternatives.