Adjunct, Three Campuses

I lost a small one purely on evidence, not on merit. We had moved to a messaging app for speed, so the version history and the approvals lived on my phone rather than anywhere official, and none of it counted. Since then everything decision shaped goes into the platform or into email even when the real conversation happens on a call, and I send a one paragraph summary after every call.

Buying a worse flat than the one you passed on is the exact outcome I am afraid of, and it is the one nobody models.

40 minutes a month is 8 hours a year. At $99 that's roughly $12 an hour to buy your time back, which is either an obvious yes or an obvious no depending on your rate and how much you secretly enjoy the spreadsheet.

Good reframe. I was treating it as "do they know the answer" when it's really "do I learn my own foot".

I did two pieces of free work for exposure in my first month, and neither client left a review because there was no transaction to review. All I got was two case studies I was not allowed to show. If you do anything below rate, make sure it still runs through the normal process end to end so it produces the artifact you actually needed.

Write review notes even when there is genuinely nothing to explain. Three lines: what the app does, the exact flow to try, and "no account, no network, works in airplane mode". It is the cheapest insurance available against a reviewer with forty apps in the queue and ninety seconds for yours.

It filters clients as well. Anyone who will not pay your rate for four hours of work was going to be difficult at forty.

I kept a vision plan for six years out of inertia and used it twice. Best guess is I spent several hundred more than the glasses would have cost me, and I never once claimed the discount portion because I forgot it existed. If you do keep it, put a calendar reminder for when the benefit resets, because unused annual allowances do not roll over on any plan I have had.

The truncate-between-tests advice is not stupid, it is just old. It comes from single-threaded suites where it was correct and cheap. It got repeated for fifteen years without the assumption attached.

I let three renewals go through untouched because the increases were small each time. Adding them up later, the premium had risen by more than half against a competitor quote for the same cover. Twenty minutes at each renewal would have been the best paid twenty minutes of my year.

Works better than it should, but you lose bids you should have won and you never learn where the time is actually going.

Everyone's is between 1.5 and 2 and everyone believes theirs is 1.1.

Three changes fixed this for me. First, quote a range with the assumptions listed underneath — content supplied by date X, one language, no ecommerce — and say plainly what happens when an assumption fails. Second, estimate in tasks and include the boring ones everyone forgets: kickoff call, revisions, testing, handover, and the four hours of admin. Third, track your actual hours on every job even when it is fixed price, because you cannot price what you have never measured. After six jobs you will have a personal multiplier. Mine turned out to be 1.7 times my gut estimate, consistently.

Then tell everyone who has ever seen your work — old colleagues, people who did not hire you six months ago, the person who ghosted you in March. That list is longer than you think and every name on it is warmer than a job board.

Twenty percent reply and ten percent close is roughly normal for cold outbound, and cold outbound is a terrible way to get work long term. The rate to watch is not proposal to close, it is where the lead came from. My close rate on referrals is somewhere around 70 percent and on cold applications it is under 10, with the same proposal text. The fix is not better proposals, it is different leads: tell every past client you have capacity, be visible in one specific niche, and stop competing in places where 40 people apply to each posting.