Every article about raising your rates hands you the same email template and the same pep talk. Almost none of them answers the question that actually matters: which clients should get which increase? Send one blanket number to your whole book and you overcharge the accounts you should be protecting while undercharging the ones quietly eating your payroll.
The fix is to reprice the way you would fix any other margin problem: measure per-client profitability first, sort clients into tiers, and give each tier its own number, its own letter, and its own notice period. Done that way, a rate increase stops being one terrifying announcement and becomes three small, defensible decisions.
Find your effective hourly rate per client
Before you touch pricing, compute one number for every client: revenue divided by actual hours delivered over the last six to twelve months. Not quoted hours, not contracted hours - the hours your team really spent, including the check-in calls, the revision rounds, and the quick favors that never hit a timesheet cleanly. That quotient is your effective hourly rate, and it is the closest thing a service firm has to per-client gross margin.
Run it across the whole book and the result is usually uncomfortable. Most owners find a roughly 3x spread between their best and worst accounts. A $6,000-a-month retainer that consumes 30 hours is earning $200 an hour. A $4,000 retainer that has quietly ballooned to 60 hours is earning $67 - same team, same overhead, a third of the margin. Until you see those two numbers side by side, both clients feel the same, which is why blanket increases fail: they preserve the spread instead of fixing it.
Sort every client into three tiers
With effective rates in a spreadsheet, sorting the book takes twenty minutes. Every client lands in one of three buckets:
- Healthy accounts: raise 3-10%. Effective rate at or near target, stable scope, good working relationship. These get the routine annual adjustment - enough to keep pace with your costs, small enough that nobody escalates it.
- Underpriced accounts: raise 10-25%. Effective rate below roughly half your target - the sub-50%-margin tier. These accounts are being subsidized by everyone else, and the increase should fix that in one step, not nurse it along at 4% a year for a decade.
- Strategic accounts: grandfather briefly. The referral engine, the anchor logo, the client with a real expansion path. Hold their current rate for a defined window - two quarters is plenty - with the end date in writing. A grandfather clause with no end date is just underpricing with extra steps.
How big an increase is normal
Calibrate against what the market already tolerates. A 3-5% annual adjustment keeps pace with costs and rarely draws more than a one-line reply; clients who manage their own vendors expect it. When you have not touched pricing in years, 10-15% is a normal catch-up, not an outrage - most clients know they have been paying last cycle's rate. The 10-25% band is reserved for genuine corrections, where the effective hourly math makes the case for you.
And the fear itself is miscalibrated. Across studies of why clients actually fire firms, price lands around sixth on the list. Weak results, poor communication, slow responsiveness, and feeling like a small fish all rank far higher. Clients leave firms that go quiet and miss deadlines at the old rate long before they leave firms that do great work at a new one.
Notice periods, effective dates, and sequencing
Mechanics matter as much as the number. Give 30-60 days of notice for most clients - enough time to adjust a budget, not so much that the increase becomes a season-long negotiation. Enterprise clients need 90 days so the new number can ride their budget cycle instead of fighting it. If a contract is in force, the increase lands at renewal, not mid-term.
Set the effective date to the first of a month, state it explicitly in the letter, and never apply it retroactively. Then sequence the sends: correction letters first, because those conversations take the longest; routine adjustments next; grandfather notes last - the easiest calls. Track replies the way you track invoices - a rate letter nobody acknowledged is a dispute waiting for the next billing run.
The three letters, word for word
Adapt the wording, keep the structure: state the change, name the effective date, give one honest reason, close without groveling. No apologies, no rising-costs boilerplate, no asking permission.
The routine annual adjustment, for healthy accounts: 'Hi [name] - a quick note as we plan the next quarter. Effective [date, 30-60 days out], the rate for your [scope] moves from $X to $Y, an adjustment of about 4%. Nothing about your team or scope changes, and no action is needed on your end - your [month] invoice will simply reflect the new rate. Happy to talk it through on our next call if useful.'
The correction letter, for underpriced accounts: 'Hi [name] - I want to flag a pricing change with plenty of notice. When we set your rate in [year], the engagement looked different from what we deliver today: [one concrete line - weekly reporting, the second campaign, same-day turnaround]. Effective [date], your rate moves from $X to $Y. That is a bigger step than a normal annual adjustment, so here is the honest version: at the current rate we cannot keep staffing your account the way it deserves. If the new number does not work for you, I understand completely and we will make any transition smooth - but we would much rather keep doing this work together at a rate that supports doing it well.'
The premium repositioning note, for strategic accounts: 'Hi [name] - some news I wanted you to hear from me directly. Over the last year we have moved the firm toward [senior-led delivery / a narrower specialty], and our pricing is moving with it - new clients now start at $Y. Because you have been with us since [year], your current rate is locked through [date, two quarters out]. After that it moves to $Z, which still sits below our new-client rate. Nothing changes about your team or scope in the meantime; I just did not want you to hear about new pricing secondhand.'
Expect a quiet yes
Most owners brace for a revolt and get a shrug. The healthy tier replies 'sounds good' or does not reply at all. The strategic tier appreciates hearing it first. The correction tier is the only place real conversations happen, and you priced those to be fine with either answer. Hold the number - a rate you negotiate down at the first pushback teaches your best clients that the letter was an opening bid.
Then put repricing on the calendar. An effective-rate review every year and a small adjustment every cycle beats a painful correction every five. The firms that never have a dramatic rate conversation are not the ones charging the least - they are the ones who never let the spread build up in the first place.
Frequently asked questions
Thirty to sixty days for most clients, and about 90 for enterprise so the new rate can be built into their budget cycle. If a contract is in force, apply the increase at renewal. Either way, name an explicit effective date and never bill retroactively.
A 3-5% annual adjustment keeps pace with costs and rarely gets questioned. If you have not raised prices in years, 10-15% is a normal catch-up. Reserve 10-25% for accounts where your effective hourly rate is far below target and you would genuinely be fine losing the client at the old price.
A few might, and on badly underpriced accounts that can be a net win. But price ranks around sixth among the reasons clients actually fire firms - weak results, poor communication, and slow responsiveness all rank higher. Clients rarely leave good work over a well-explained increase.
No - grandfather deliberately and briefly. Give strategic accounts a written window, typically one or two quarters, then move them to the new rate, ideally still slightly below what new clients pay. An open-ended grandfather clause just recreates the underpricing problem you set out to fix.