The Problems of Working with Web Agencies
Post from December 11, 2014 (↻ May 18, 2018), filed under Everything Else.
This and many other posts are also available as a pretty, well-behaved e-book: On Web Development.
I started my career in a small agency, I later worked for a big agency, and I at other times collaborated with or managed agency staff. I’ve never enjoyed working for nor with agencies. That was not because of the people, but because of some inherent issues with how agencies work: a modus operandi that has always been opposite to my ideas of commitment and quality.
In two brief articles I’ll lay out my perspective on web agencies: here, what the issues are when working with them; later, about ground rules when we still end up working with them, to get the best out of the relationship. Much of what follows is inspired by an investigation I led at Google, one that yielded similar conclusions.
For a definition, I simplify and understand a “web agency” to be either a third party that designs and creates websites or apps, or a branch of a third party that does so.
- Different Interests
- Poor Knowledge of Company and Users
- Wide Quality Range
- High Cost
- No Maintenance
1. Different Interests
The biggest issue with agencies is that they don’t have the same interests. The agency looks at revenue and reputation through their clients. While short-sighted, the current job may be enough to bring in more money and get another reputable client on the portfolio. Even with the most benign and professional of agencies, interests can’t be aligned—otherwise they were competitors.
2. Poor Knowledge of Company and Users
Agencies have little if any knowledge of the client and their users. One can say that the better the agency, the more they know or try to know, but here, too, there’s a logical boundary in that an agency rarely knows as much as their client. And that makes them already less useful to cater for company and user needs, and represent these well.
3. Wide Quality Range
We must be careful for there are a good number of agencies that deliver high quality (and I believe at Aperto we had usually improved what clients had), but generally there are big differences in quality. We can shield ourselves from poor quality in the scouting and procurement process—if we do our homework well—, however there’s even with good agencies a risk of subpar work.
To counter the softness of this argument, agencies come with several risks, like the risk of security issues as well as the risk of leaking information. Both happens inevitably again because we’re dealing with third parties.
Agencies then mean inconsistencies in branding and coding. This happens mostly because of poor knowledge, now of existing client infrastructure—the web developers of us know how it is to on-board an agency on our web frameworks—, but also again when agencies change. When a company’s sites, apps, ads look all a little differently in branding aspects it’s probably because of one or more agencies’ involvement rather than because of inattention on behalf of company staff.
5. High Cost
Agencies also mean significantly higher cost than inhouse work. One estimate I have in mind is by a factor of 10; then I recall projects that could have been done inhouse in two weeks by one person, but were instead delegated to an agency that charged the equivalent of a yearly salary. Comparatively speaking and taking all factors into account, agencies are extremely costly.
6. No Maintenance
Last but not least, agencies don’t maintain their work. That means a deferred tax for what agencies do needs to be maintained, too (even if taken down). From my experience, maintenance gets terribly neglected in agency projects—and although so is maintainability in web development in general, agencies shouldn’t add to the already unnecessarily high maintenance load of most companies.
❧ What are the benefits, then? Agencies help us deal with one-off needs and excess load, and they can fuel us with an outside perspective and fresh creativity. Whether we can attain this, in the light of all the cons, only through agencies and not through, say, smarter hiring, I don’t want to discuss here. But there are some things we can do when working with agencies, which I’ll share in the next post .
About the Author
Jens Oliver Meiert is a tech lead and author (sum.cumo, W3C, O’Reilly). He loves to try things, particularly in the realms of philosophy, art, and adventure. Here on meiert.com he shares and generalizes and exaggerates some of his thoughts and experiences.
If you have any thoughts or questions (or recommendations) about what he writes, leave a comment or a message.
Before I continue, a quick disclaimer - I’ve never worked for an agency - always internal but have engaged many agencies in my time.
I’m not sure the differing interests is a fair criticism in the way you’ve phrased it Jens.
While an agency does have revenue as a goal, it is a reasonable goal - they are a business after all. Every business you’ve worked for in your life have revenue as a goal, it is like air - you die without it.
I have no doubt that agencies use their clients reputation as a springboard to help them gain the next client but again, that isn’t unique to an agency - it is common in business generally.
A point I’d hope would keep agencies honest is that while they are chasing revenue (okay) and reputation via association (okay) - if their work quality isn’t up to to scratch, it is the job of the next incoming client to complete due diligence to weed out the crap agencies by contacting their past clients (for example).
In practice this may not happen anywhere near enough but if it was more widespread, agencies would be in a real reputation battle to win new clients - as the expression says - their reputation would literally precede them.
The remaining points I can’t really argue against, I’ve seen each of those in action in some form from an agency in the past. Frustratingly, in spite of weeding out low quality agencies - even top notch agencies drop the ball from time to time and the client gets a lemon.
This is also true of internal teams, from time to time, despite every intention and process to deliver rock star outcomes - it just doesn’t hit the mark and it takes more time, money or resource to rectify the lack poor showing.
On December 11, 2014, 15:00 CET, Bruno Baketaric said:
let me open another perspective.
This largely depends on the relationship that is built between the client and the agency. If the client just want’s to “outsource a Job” - for whatever reason - you may be right: different interests, short term cash flow.
If the relationship is built around the terms “partner” and “collaboration” things change: in that case, the agencies interests (should) change to make the client “successful” - whatever that stands for in a particular case.
Poor Knowledge of Company and Users
Of course there’s a boundary (first), but in a long term collaborative approach (from both sides), this boundary blurs and vanishes. It can even turn around: I know cases, where employees on the agency side are working THAT LONG for a client, that they know e.g. the Products better than marketing employees at the client.
Wide Quality range
While true in general, quality is more often limited by time and/or budget. See Triple Constraint.
Only true if you do “Agency hopping” or if you hire an agency that lacks knowledge in branding.
While I’d say your example of a factor of 10 is way too high and unusual, I generally agree. On the other hand you should also take into account that the Client trades Money for building up internal Resources and internal Knowledge. Of course a company like Google doesn’t have to “buy” Knowledge in Web-related stuff (I assume), but for most of the companies it’s not feasible to build up a amount of Web-Knowledge and Experience with internal Resources that can be “bought” from a Web-Agency - and to keep that Knowledge updated, which is quite hard nowadays (think of AngularJs, SASS/LESS, the “Multi-Device-World” etc. some years ago).
Again, depends on the relationship. I’m sure no agency will (deliberately) deliver hardly maintainable Code (or whatever the output is), if they know, that they will be held responsible for that in a long-term client relation.
The true problem here is “agency hopping”, passing individual “Jobs” (without giving the agency a chance to see the big picture) to that particular agency that submitted the cheapest offer. In that case, your concerns are right.
But they are all wrong, if the client seeks for a long-term collaborative partnership with an Agency - and most agencies are willing to build up that kind of relation.