How Do You Set Reasonable Marketing Benchmarks?

You have to know what to aim for when you start marketing. So a lot of people, when realizing this fact, try to find a “reasonable” marketing benchmark.

That number, of course, is just one pulled from a chart. The average conversion rate, the typical cost per lead, the good email open rate.

But numbers without context don’t mean much. Average for what industry? Typical for what kind of lead? Good email open rate for whom?

Numbers from charts feel authoritative. But benchmarks copied off charts are the fastest way to set a target that is trivially easy or quietly impossible. And you won’t know which until you’ve spent a lot of time, and probably money, chasing it.

But you still need marketing benchmarks. So how do you set them?

My answer: you use your own economics and your own data. When you don’t have data yet, you come up with a decent, thoughtful, and clearly labeled guess that you’ve committed to replacing the minute you have something more solid.

I started down this road in an earlier piece on knowing your marketing is working, where I said to use industry standards at first and slowly accumulate the data you need to make your own. This post is me diving more into how to do that. Namely, I want to talk about how to set a bar that's high enough to mean something, low enough to be real, and grounded in something other than a stranger's spreadsheet.

A note on scope before I go on: this post is about threshold-setting. It’s about how high the bar should be and where it comes from. It is not about which metrics to track. That's a different question, and I've written about it separately, so I won’t retread the same ground here.

Benchmarks and key performance indicators (KPIs) are used like synonyms, but they should not be. Benchmarks are external reference points based on what others are doing, such as the industry as a whole. A KPI is a measure of your progress toward your goal. They do different jobs.

I point this out as a matter of hygiene early on. In this context, numbers don’t mean anything without something to compare them against, like a target, a prior period, a forecast, or a benchmark. Your benchmarks exist because they give your KPIs context and help you set a target. But the KPI is not the target itself.

Benchmarks help you get a lay of the land. KPIs tell you whether you're moving toward your goal.

Now let me make the case against the thing you came here for: the industry average.

The core problem is that averages are wide and they hide enormous variation. This is the same trap I've flagged with content marketing costs and with SEO tool data: a single "average" smears together businesses, verticals, intents, devices, and offers that have nothing to do with each other. An "average conversion rate" across an industry is built from companies you'd never compare yourself to in a million years.

The honest way to treat any industry figure is the way you should treat any model. It’s an estimate with an unknown confidence interval, but it’s not ground truth. As the old saying goes, all models are wrong, but some are useful. The trick is never forgetting you're using a model.

But the single most damaging way benchmarks go wrong isn't the averages. It's the arbitrary target. A founder or an exec declares a number because the company "should" be hitting it. It feels motivating in the moment. But if the bar is set at an arbitrary or unreachable height, it throws the team into a tailspin. It's demoralizing, and demoralized teams perform worse. A reasonable benchmark is grounded, not aspirational-by-decree.

That’s why I’m a fan of treating benchmarks like diagnostics not goals. They’re best used to spot where you lag the field. Then you have a chance to go investigate why. Don't adopt the average as your target just because it's the number you happened to find.

If you can’t just use the industry average, then what can you use? I suggest you use two anchors here to help you set your own benchmarks.

Your unit economics set the ceiling. What you can afford a metric to be is a hard, defensible bar that no industry average can hand you. The clearest example is cost per lead: the most you can pay before you lose money comes straight from your own margins and close rate. It’s your gross profit per customer times your close rate. That's the same break-even logic I use for budgeting Google Ads, and it's the bar that matters when you’re talking to your accountant about it, because it's tied to whether you make money.

Your own history sets the trend. Once you have data, your previous period is the benchmark that matters most. Am I improving against myself? That question is almost always more useful than "am I beating a national average built from companies I've never met?"

Here, I’ll suggest a couple of refinements to help keep your benchmarks helpful and honest. First, segment by source and intent—an SEO lead, a cold-email lead, and a paid lead convert at completely different rates, so a single blended benchmark hides more than it reveals. (This is exactly why comparing conversion rates across traffic sources is where the wheels fall off.)

Second, judge against downstream quality. A higher conversion rate isn't automatically better. If it came from a flood of unqualified leads, you've made your numbers look great and your business worse.

The larger principle at work here is that a reasonable benchmark is built from the inside out. Use your economics first, your own data second, and industry figures only as loose context.

Suppose you have no history. The listicles never seem to talk about this problem, but a lot of people run into it. If you have a brand-new business or you’re starting a new channel, the slate is blank. You can’t benchmark against your own data because you just don’t have any yet.

It’s here that the best thing you can do is set an explicit, low-confidence prior. You can borrow it from an industry figure or make an educated guess. But no matter how you source your prior, you have to label it as a guess. It’s not truth. You commit to replacing it as soon as you have better data.

This is exactly what a good test plan does with its working assumptions: you write down your best estimate, you flag it as low-confidence, and you update hard as evidence comes in.

In practice, this tends to break into five steps:

If you want the jargon, this is just Bayesian common sense. You start with your best guess, hold it loosely, and update aggressively as evidence arrives. The big mistake is not using an industry number to get started—it’s forgetting you made a guess and failing to fix it with better data when you have it.

I’ve already covered a lot of ground here, so I’ll summarize the framework in six steps. You can use these steps for any metric you want to track:

It’s tempting to skip that first step because the benchmark would appear to be “self-evidently helpful.” But don’t do that, because writing down which benchmark drives which decision is what keeps your dashboard a decision tool instead of a wall of numbers you glance at and feel vaguely good or bad about. (I've made the case for tying everything back to jobs booked and revenue rather than vanity metrics if you want the fuller argument.)

Frameworks are easy to nod along to and hard to apply, so here are three quick examples. Each follows the same basic pattern of borrowing a benchmark, then building a better one.

An SEO benchmark. The trap is benchmarking rankings or raw traffic. These are classic vanity metrics. The reasonable bar is leads and revenue, judged as a trend over a realistic window, because SEO compounds on a delay. Borrow a rough timeline as your prior ("results often take six to nine months to really show"), then replace it with your own ramp data as it comes in.

A Google Ads benchmark. Don't benchmark your cost per click against the industry. Benchmark your cost per qualified lead against your break-even economics. An industry CPL is a loose prior at best. Your own math sets the ceiling that actually determines whether you're winning or losing.

An email benchmark. Here's a usable starting prior: I tend to treat a 40% open rate as a rough rule of thumb, with the caveat that open rates are a directional early-warning metric, not the goal. Clicks and revenue matter far more anyway. And know that these vary enormously by industry; click-through norms in, say, legal look nothing like eCommerce. So the borrowed number is always provisional. It gets you moving, but it doesn't get the final say.

Notice that with all three, the industry figure is where you start. Your own data is where you land. And if you're wondering which metric to even watch for a given channel, that's the metric-selection question—this post is about how high to set the bar once you've picked one.

A benchmark you can't defend from your own numbers is just a guess you've committed to. So with benchmarking, the real work isn't finding the right average. It's knowing your economics and instrumenting your marketing so you can outgrow the borrowed numbers as fast as possible.

Start with a labeled guess if you have to. There’s nothing wrong with that. It’s often a lot better than nothing.

But hold your guesses loosely and watch the trends. Then replace them with your own data the second you have enough. A reasonable benchmark is one you can defend by tying it back to the bigger business goals of revenue and profit. Everything else is provisional.

My company helps B2B service businesses generate qualified leads through data-driven SEO. We do the work and we build the tracking to show you what's producing results.

If you're interested, book 30 minutes of my time and we can talk about whether it makes sense for your business.