I’m speaking as someone who has been a Director of a 30-strong production team which included project managers, IA/UX, graphic designers, and front+back end developers at an agency. Caveat: I haven’t worked for many agencies but have just noticed that the models are similar; and what I’m saying may be local to Montreal. A big part of my job was to vet estimates that go out of the door to make sure they are realistic; and it was also part of my work to make sure the right people with the right skills/inclinations got on the right projects. I was also involved in projects in very early stages to help vet the technology solutions (or assign someone to do the same), and make sure people talk to who they need to talk to at concept stage.

It took about 2 years to change the company so that collaboration happened and tech took its place at the strategy table rather just the bit that you shove in at the end to make it work. That’s probably another conversation to be had.

There are many things wrong with agencies. Some of it is timescale and estimates, as you mentioned, a lot of it is cultural and directly related to how agencies run and make money. The following are my own observations:

1) Agencies often charge by a day rate or an hourly rate. This often means that almost everything needs to be billable. It also means every time you have a team meeting it costs money per team member. The nature by which agencies charge means it’s not economical (on paper) to have too many people involved in the project because it’s hard to make that money back. If you do, you end up charging a big overhead that makes you less competitive with other agencies.

2) Agencies under pitch in order to win. Unfortunately, we are a market that competes first for price rather than for quality. I used to get into fights with account managers who under pitch because they want to win the client. The pitch is meant to seduce. Production is not, strictly speaking, sexy. The result is that the budget is way too low for the requirements that the clients want.

3) Budget estimations at concept are never revisited prior to production. This is one of the seriously dumb mistakes that everyone makes. The pitch concept/design to seduce cannot be the thing that gets produced (there’s not enough information to go to prod), but no one ever has the guts to revise the cost and timeframe between concept and production, even after say, detailed wireframes have been produced — if it was 😉

Before anyone jumps forward, I want to say that the “agile method” is not an answer. Agile is great for developing products, especially with dedicated teams. But if you’re running short-term projects with the same people across projects, you need something hybrid that’s going to map directly to how the company earn their daily bread.