The same could be said for IT and software development but for a strange reason lots of people are doing it.
The same could be said for IT and software development but for a strange reason lots of people are doing it.
I think for software it's easier to specify requirements that more closely relate to goals. Then you pay for the complexity they entail. For instance, you can say I want to replace this form with the objective of eliminating paper and improving efficiency. The dev shop might charge you more than the market rate but they'll still have to deliver you something that's relatively close to what you imagined.
Furthermore, it would be in the interest of an honest dev shop to advise and build something really useful for the client so that they come back.
I'm struggling to understand what are the differences - why isn't the best interest of a marketing agency to generate good results to a business? What's not aligned in that?
Like... it's cheaper to find clients, fool them into getting a chunk of their marketing budget, and dump them? I don't see how this is good business.
Why can't a dev shop just do a bad job, hide behind poor requirements, and keep a infinite scope of work after you're too invested in them?
The thing with consultancy it's kind of what you've said: they've built models/systems that kind of work with minor tweaks and sell it to other companies.
It's done because the perception of building those systems from the ground up seem more expensive then to get it "installed"... which might be true for many scenarios, but for others, it might not.
It's that specifying the meaning of a "good job" is much more difficult when it comes to ads.
Good job in software means do what I want and do a great job means accomplish what I want but do it more efficiently. (And on budget ofc)
For ads do a good job means give me most customers for my budget. And the measure of that is CAC. But I've yet to come across an ad agency that bases it's compensation on CAC.
Well that's the crux of it: it's about people and people aren't input/output machines. One message (input) doesn't predict a behavior (output).
That's why I think that anyone that promises specific results in marketing, you should step away from it. I think the best stance is that they will work with you to achieve something, and "work with you" means literally that - try to figure things out with you because your brand/product is it's own thing.
>For ads do a good job means give me most customers for my budget. And the measure of that is CAC. But I've yet to come across an ad agency that bases it's compensation on CAC.
Ok, now I think I understand you...
From the moment you're dependent on a 3rd party, basing your compensation on their work it's just soaking up risk. This goes for both ways: it's the same of some marketing agency saying to a dev shop hey guys, our marketing is solid, don't worry, we will pay you in rev share until the bill is settled... why would you take this deal?
Same goes for marketing: how can they know that you have a good sales dpt. to close leads? How can they know you have a good on boarding? Is the price competitive? They could be throwing you the best ads, getting top notch potential customers, and your product just sucks. Now what? Why should the agency pay for your bad job?
BUT they can establish a baseline CAC over the first couple of months and then work on improving it.
Take a look at this comment as well: https://news.ycombinator.com/item?id=27645674
I've spent years at the intersection of dev work, marketing, advertising, analytics, SEO, etc. for small businesses. Now the I know how terrible a job they usually do, I'd never pay a consultancy to do that work. I probably wouldn't even bother with it myself. It's mostly just faith based prayers to a few ad marketplaces who's interests are inherently in conflict with yours. It's a shitty marketplace all the way around.
Dev work has a higher barrier to entry, for now, so it's less easy to bullshit your way through with clients. The resulting software either does what it's supposed to (with bugs, of course) or not. Whereas the successful or failure of many digital marketing campaigns are largely matters of faith and not evidence.
That's why it's easy to perceive that marketing is mostly shit... because it is, and no wonder, it's not "hidden" in the backend or in some git repo.
Ads live in public scrutiny.
A lot of the dev work would probably just make you cringe just like many ads. You just don't see it.
With that said, you're right that the barrier of entry it's lower, and I don't think the hard part of marketing/advertising is the campaign setups - that's mostly being reduced to a step-by-step wizard. Thankfully! But that won't make marketing any good, in fact, you'll just see more of the bad stuff.
Marketing work is never ending and very fluid. There is always more you could be doing and you could always be doing something better with more time. It inherently doesn’t work well as spec work. And yet that is how marketing agencies operate.
Agencies are best hired when you do need something very specific done that you don’t have the expertise to do in-house. A video, a new website, a very specific, time-limited marketing campaign.
If the work is continuous, it should be moved in-house ASAP and it is better to wait then pay an agency to do it in the short-term.