I would guess a repeatable/known type of service or project helps a lot as well. Most of my engagements are relatively different from each other, and I'm doing different things all the time. I've worked places which were offering a narrower set of functionality, with limited customizations, and that's generally much easier to estimate. Also much easier to have estimates in which you've said 'no' to things altogether, vs spending time researching/testing/trialing stuff that's new to you.
To the point of providing flat rate pricing at my job. We realized spending an hour or two getting a sign off on an SOW or spending four hours on a discovery call didn't change the time it took to complete in 90% of cases. I can just price my stuff all the same because for the most part, EDI mapping is EDI mapping. If we run into something weird. Yeah. But thats once or twice per quarter.
Meanwhile our ecommerce integration product pricing is all over the place, because there is just so much variation. I am hoping to move that to base pricing for the product + time and materials for consulting.
In a sense my EDI team is busier, less stressed, but less gratified. The web team is always getting a curve ball and its certainly a younger man's game. Or at least someone who's willing to roll with the punches, but the payoff is sweet
Well for me it certainly helps that I usually have clear requirements up-front. Designers for instance take a lot of the guesswork out of what I need to do for website work. For other things, like debugging legacy code, I just tell clients that I need time to diagnose the issue before I can give an estimate.
