The 90-Day ROI Dashboard Plan: One Baseline, One Workflow Fix, One Outcome

Pick one baseline, one workflow fix, and one outcome. That's the whole quarter.

The most common reporting failure at a law firm isn't bad data. It's too many simultaneous priorities. A firm decides this is the quarter it finally gets serious about measurement, and then it tries to rebuild attribution, replace the intake software, test three new channels, and add an AI visibility tracker — all at once, all in ninety days. At the end of it there are more dashboards than there were in July and not one decision that got made differently.

The fix is structural. Give each month of the quarter exactly one job. Month one: establish a baseline. Month two: fix one workflow. Month three: verify one outcome. Nothing else gets added until the ninety days are done.

Why response time is the workflow fix worth choosing

If you only get one workflow repair per quarter, speed to first contact is almost always the highest-value candidate — and firms know it. Thirty-five percent of firms in CallRail's 2026 survey estimated that slow response had cost them 11%–25% of annual revenue.

Sit with that range for a second. Not 11%–25% of leads. Of annual revenue. That's a number large enough that it dwarfs almost any media spend decision you'd make in the same quarter. You can win a percentage point of conversion by rewriting a landing page. You can't win a quarter of your revenue back that way.

It also explains why so many marketing audits come back inconclusive. When response time is leaking that much value, every channel looks mediocre. The paid search numbers look soft. The referral numbers look soft. Organic looks soft. You cannot diagnose channel quality through a broken intake layer, because the intake layer is applying the same tax to all of it.

You cannot diagnose channel quality through a broken intake layer — it applies the same tax to every source, and makes all of them look equally mediocre.

Month one: measure payback by campaign

Before you change anything, you need a number that isn't a vanity number. The baseline for month one is payback by campaign — not leads, not cost per lead, not impressions. What did each campaign cost, and what did it return in signed matters?

The discipline in month one is resisting the urge to act on what you see. You will find a campaign that looks terrible. Leave it running. You're establishing a pre-change reading, and if you start tuning spend in week two you'll never be able to separate the effect of the workflow fix from the effect of your budget shuffling.

Practical scope for the first thirty days:

• Every active campaign gets a spend figure and a signed-matter figure for the same period.
• Define one consistent lookback window and apply it everywhere. Firms in this situation often let each channel report on its own timeline, which makes the comparison meaningless.
• Write down the current time-to-first-contact, even roughly. That's your before-picture for month two.
• Note which practice areas and cities are producing the payback, not just which campaigns.

Month two: fix response, and report counts and rates

Month two is where you attack the leak. But the reporting rule for this month matters as much as the fix itself: show both counts and conversion rates, or your comparison will mislead you.

Here's the trap. A campaign that produces 400 inquiries at a 6% signed rate and one that produces 60 inquiries at a 22% signed rate look completely different depending on which column you show. Count-only reporting makes the high-volume campaign look like the winner and pushes budget toward the cheapest possible traffic. Rate-only reporting makes tiny, high-intent sources look infinitely scalable when they aren't. Firms make expensive budget decisions on this distinction all the time, in both directions.

Put the two side by side in every view. If a channel improves its rate while its count collapses, you want to see that in the same glance — because that's usually a targeting change, not a quality win.

Meanwhile, the workflow fix itself is narrow on purpose: shorten the gap between a prospect reaching out and a human responding. Route after-hours contacts somewhere real. Kill the handoffs that add hours. Assign ownership so no inquiry sits waiting on "whoever picks it up."

Month three: verify the effect on AI citation trend

Month three is verification, and the outcome worth verifying now is AI citation trend — whether your firm is being surfaced and cited when people ask AI search tools for legal help in your markets.

This is the layer most firms are still measuring by anecdote: someone at the firm asks a chatbot a question, doesn't see the firm, and panics. That's not a trend. A trend is a repeated, consistent check over time in the queries and cities that matter to you, watched over a quarter.

Coverage is the input variable here. AI systems cite pages that exist and are structured to be understood. A firm with a handful of general practice-area pages has very little surface area to be cited from. That's the logic behind our Bosseo 10k approach: roughly 10,000 service-and-city pages on the firm's own domain, covering every practice area across every city served, with schema, internal linking, and AI-search optimisation built in. It's not a replacement for an existing agency and it isn't "better SEO" — it's geographic coverage infrastructure, the page inventory a firm will never build by hand. Typical engagements run $1,750–$5,997/mo depending on state and city count, often with a build fee, and it fits established firms already ranking somewhere who want to own more cities and practice areas.

Why late summer is the right window

Timing is not incidental. Late summer is the useful moment to fix the systems that will carry Q4 demand. Whatever your intake process looks like in November is what it looked like in August — nobody rebuilds workflow during their busiest stretch. Do it now and Q4 volume lands on a system that can hold it. Skip it and Q4 volume lands on the same leak, at a higher cost per inquiry.

The distraction clause

Write one more line into the plan: no new tools for ninety days. A quarterly plan is stronger when it explicitly protects focus from new-tool distraction. Something will get pitched to you in week five that promises to solve everything you're currently working on. It won't, and evaluating it will cost you the month.

Park every new tool in a list. Review the list on day 91, when you have a baseline, a fixed workflow, and a verified outcome to judge it against.

Your next step

Before the week ends, do the smallest version of month one: open your last full month of spend and pull signed matters against it, by campaign, on one consistent window. If you can't assemble that in an afternoon, that gap is your month-one project — and it's worth more than any campaign you could launch instead.

If you'd rather not build the reporting layer yourself, see how Bosseo approaches the ROI Dashboard — payback by campaign, counts and rates together, and AI citation trend in one place.