Added bookkeeping to the service menu, closed it eight months later down $9,400

The flip job costing thing is not a bookkeeping problem you can solve on your end. I've handed three different bookkeepers a rehab and every one of them wanted invoices coded to a cost bucket, and my crew sends photos of receipts with no job written on them. Until the operator fixes intake, you're guessing, and guessing at 3.5 hours a month becomes nine hours a month. You were right to drop those.

The loss is almost entirely in one line. Six clients at 19 average onboarding hours instead of 4 is 90 unbilled hours. At the rate your steady state implies, call it $79, that alone is about $7,100 of the $9,400. Fix the cleanup pricing and this line is roughly breakeven in month eight even with the year end collision.

The assumption that did the damage is that historical cleanup hours are predictable enough to fold into a flat monthly. They aren't, because the variance is driven by things you can't see before you have the file open: commingled personal spending, prior bookkeeper artifacts, property manager statements entered at both summary and detail level, unreconciled periods of unknown length. Cleanup is normally scoped as a diagnostic first, a couple of paid hours where you open the file and count the specific problems, and only then quote the project. That converts an unbounded risk into a small one you get paid for either way.

The second thing worth changing is the pricing unit. A flat monthly per client prices the relationship. The work scales with transaction count, number of bank and card accounts, number of entities, and number of properties, and those four move independently of each other. A landlord with two doors and one account and a landlord with two doors, three entities and a construction line are not the same monthly job at any price.

On the year end collision, that's a capacity question rather than a pricing one. If your other line peaks in the same eight weeks, the bookkeeping client count you can carry is set by that window, and you sized to your quiet months instead.

What happens to the math depends on whether those cleanup hours were genuinely one time. If clients keep reopening prior periods, steady state was never 3.5 hours and the whole model needs re-testing.

I've been the 41 hour file. My former bookkeeper never told me the duplicated property manager deposits were an issue, she just absorbed it and then raised my rate 60 percent with no explanation and I left, annoyed. If she'd shown me the mess and charged me to fix it I'd have paid and stayed. People underestimate how much clients will accept when they see the invoice line.

$79 an hour at steady state seems okay to me but I have no reference point. Is that normal for this work or is that low?

@plumb it depends on whether you're selling hours or a monthly deliverable, and on what you're on the hook for. The effective rate matters less than whether the client can generate unlimited questions inside the fee. That's what got @beacon. Five years in, my rule is that anything with an open ended question count gets a scope boundary written into the agreement, or the good clients subsidize the one who calls on Sundays.

Also @beacon, closing the line and moving four clients to a firm is the right ending and most people don't do it. They limp along at negative margin because canceling feels like failing.

Harrow's point about receipt intake is the whole thing. On the trades side I've never once seen a receipt come back with a job number on it unless someone made it painful not to.

I'm mid build on a service business and this thread just changed my onboarding. The diagnostic hour that @rena-expert described, charged, before any quote, is going in. What I don't know is how to sell it. "Pay me $300 to tell you how bad it is" is a hard opening line to a prospect who thinks their books are fine.

@lantern call it a books review and give them a written findings list they own regardless of whether they hire you. People pay for a report. They resist paying for a look.

The duplicated property manager deposits happen constantly and I'll defend my side of it slightly. We send an owner statement that shows gross rent, our fee, maintenance, and a net disbursement. If the bookkeeper enters the statement detail and also the bank deposit, that's double. If they enter only the deposit, the client's P&L shows net rent and no expenses, which makes the property look like it has no costs. Neither is obviously right to someone who hasn't done it, so we now send a one page note explaining which way to book it. Cut our support emails noticeably.

@beacon, what did your client agreement say about prior period adjustments? If it's silent, every reopened month is free work by default and you'll do it forever. One clause fixes that: prior periods are closed after client approval and any reopening is billed at hourly. It's three lines and it would have paid for itself twice in your eight months.

The part I keep rereading is turning down $6k elsewhere to keep the books moving. That's real money and it doesn't show up in most people's accounting of a failed service line.

Running a small team, the version of this mistake I made was different in shape and identical in cause. I let the onboarding work sit with me instead of building a checklist someone else could run, so my capacity was my own calendar and my calendar had a season. @beacon, if you ever restart this, write the cleanup process down as steps before you sign client one, even if you're the only one running it. The document is what lets you hand it off later, and the moment you can hand it off, the year end collision stops being fatal.

Also, six clients was too many to learn on. Two would have shown you the 19 hours at a quarter of the cost.