Every adviser knows marketing review is slow. Very few can say what it costs, which makes it hard to justify changing anything about it.

The number is usually larger than expected, and most of it isn't the part people focus on. Reviewer hours are the visible cost. The bigger one is the delay between a piece of content being ready and being live.

Here's how to work out your own figure.

The four costs

1. Reviewer time

The straightforward one. Take the number of pieces reviewed per month, multiply by average hours per review, multiply by the loaded hourly cost of whoever performs it — salary plus benefits and overhead, not base salary.

Include everything that goes through review: social posts, email campaigns, one-pagers, pitch decks, website updates, third-party content. Firms consistently undercount here, because the small items don't feel like reviews.

2. Revision cycles

A piece rarely clears on the first pass. Each round costs reviewer time again, plus marketing's time to make changes, plus the context-switching on both sides.

This is the cost most sensitive to how the review is delivered. A review that returns “this doesn't work” produces more rounds than one that returns specific language with the reason attached. If your average is three rounds and it could be two, that's a third of the reviewer cost and a third of the delay.

3. Time to publish

The one that actually matters commercially, and the one nobody measures. Count calendar days from draft-ready to published.

For timely content the cost isn't inconvenience, it's relevance. Market commentary that clears review nine days after the event it commented on has lost most of its value. Campaigns that miss their window get shortened or dropped. Neither shows up in a compliance budget, but both are real.

4. Risk

Harder to quantify honestly, and worth being careful about. Two failure modes, pulling in opposite directions:

  • Things that shouldn't have gone out. A marketing rule finding at examination, with the remediation and re-papering that follows.
  • Things that never went out. Content abandoned because review would take too long. Invisible, and in a growing firm often the larger number.

You can't assign a credible dollar figure to either. What you can do is note them alongside the arithmetic rather than folding a made-up number into it.

Putting it together

The base calculation is simple:

Monthly reviewer cost = pieces per month × hours per piece × review rounds × loaded hourly rate

Run it once with your current numbers. Then run it again with the assumptions you'd expect from a better process — fewer hours per first-pass review, and one fewer revision round because the feedback is specific enough to act on. The gap between the two is what you're deciding whether to pursue.

Two rules for keeping the exercise honest. Use your own numbers, not a vendor's benchmark. And be conservative on the improvement assumption — if the case only works at an aggressive number, it doesn't work.

What automation changes, and what it doesn't

Being direct about this matters, because the category has a history of overclaiming.

What it changes: the first pass. A system can check a piece against the rule set, flag what's likely to be a problem, cite the provision, and propose specific replacement language. That compresses reviewer time on routine material and cuts revision rounds, because the writer receives something actionable instead of a rejection.

What it doesn't change: who is responsible. The judgment calls — whether a particular claim is fair and balanced in context, whether a disclosure is adequate for a specific audience — remain with the compliance officer, and should. Automation that finalizes its own review is a control failure, not an efficiency gain. Anything you adopt should end with a human sign-off and leave an audit trail showing who approved what, and when.

The first pass gets faster and more specific, the reviewer spends their time on the material that genuinely needs judgment, and the record of the decision is better than it was.

Where to start

Before evaluating any tool, measure what you have. For one month, log every piece that goes through review: date submitted, date approved, hours spent, revision rounds.

A month of real data tells you whether you have a problem worth solving, and it's the only thing that makes a vendor's claims testable.

That measurement is also the case we'd want you to hold us to.