I'm not always a fan of Guy Kawasaki's work, but really enjoyed his post on the OPEN Forum - A Dozen Don'ts for Entrepreneurs. I thought I'd take a stab at replicating it with some of my biggest warnings for those in our field.

For the list below, the word "clients" is interchangeable with "marketing manager" or "executive team" for in-house SEOs.

1. Don't Create False Expectations
   Clients are just like everyone else - when you exceed their expectations, they love you. When you disappoint, they're angry. Make it easy for yourself and don't oversell. If anything, undersell your abilities to do great things and let them be surprised. It's a hard thing to do, particularly in a competitive bidding environment, but humility and hard work often shine through in presentations and good clients will see that and honor it.
   _
2. Don't Ignore Analytics
   Website analytics, both visitor traffic and third party metrics, are important parts of SEO. When things are going well, even if best practices aren't being followed, it can be wise to match up data and trends to see what's made a real difference. Don't undertake an SEO project unless you have at least the essential data points (this also comes in handy once changes have been implemented and your work starts to have an impact).
   _
3. Don't Always Take Your Client at Their Word
   If you talk to lots of clients, you'll find that none of them have ever spammed the engines, bought a link, accidentally cloaked for Googlebot or hidden text, yet the statististics tell another story. Never assume your clients are being dishonest, but always watch out for activities they might not be aware of (or might not have realized were problematic). This goes beyond just white and black hat - we had a client who thought they had a couple dozen active domains; turns out they had nearly a hundred - canonicalization alone has been a big project and a big return.
   _
4. Don't Get Into Projects with People You Don't Like
   If ever you get a "funny feeling" about a client, move on if you can possibly afford it. Some people just don't click together, and when interpersonal relationships aren't working, projects have a way of not working out, either. It's always better to get out before something's signed than after.
   _
5. Don't Give an Unqualified Answer Unless You're Extremely Certain You're Right
   If you've been reading SEOmoz lately or hearing me speak at conferences, you'll notice that my advice comes with a lot more caveats than it used to. It's been a tough lesson, but there's very rarely a "this is ALWAYS better than that" in the field of SEO. Exceptions abound, so cage your language accordingly.
   _
6. Don't Confuse SEO & Sales
   If your client comes to you wanting to drive sales with SEO, make sure they're keenly aware of the multiple responsiblities inherent in such a request. Yes - SEO can drive lots of high quality, targeted traffic at the perfect moment for capturing the sale. But NO - SEO cannot convert that visit into dollars. If the website sucks at turning visitors into leads, do the right thing and recommend CRO (Conversion Rate Optimization) before they dive into SEO.
   _
7. Don't Rest on Your Laurels
   If you're not paying attention in the SEO world, even for just a few weeks, you can miss massive changes. Look at June! We've had a reversal of position on nofollow and Javascript links from Google, a new engine/algorithm/brand from Microsoft, adoption of rich text formatting in the SERPs, evidence that header tags may not be as valuable as we thought and data suggesting that alt attributes are highly correlated with good rankings. Stay ahead of the curve and devote some resources to industry news - you owe it to your clients and yourselves.
   _
8. Don't Undervalue Your Work
   SEO is hard work. For every consulting hour, there's days of research, testing, reading, surfing and experimenting. Don't undersell your services or accept that what you do doesn't provide tremendous value. If you're being undervalued now, consider how terrificly trackable SEO really is and show them the data. It's almost always on your side.
   _
9. Don't Believe Everything You Read
   Yes, even here at SEOmoz! We certainly try our best to provide high quality, accurate information, as do many other great sites on SEO, but no one is right 100% of the time, and, more importantly, not every piece of advice is applicable for every business or every situation.
   _
10. Don't Underestimate Dev Contributions
    I was recently asked "what's the biggest roadblock to SEO," and didn't need to think for 10 seconds before quoting Mr. Ballmer's infamous adage "Developers! Developers! Developers!" If you get bandwidth cycles for SEO projects, use them wisely. If the developers have made critical SEO errors, don't be quick to criticize - you'll make enemies, and, oftentimes, be guilty of hypocrisy. Stay humble, prioritize the big pieces and make sure you have the resources before you commit to improving traffic.
    _
11. Don't Overstate Your Influence or Abilities
    Just because you have the ear of some important minds at Google/Yahoo!/Facebook/etc. doesn't mean you can influence change within these large organizations. I've heard a lot of stories from companies that worked with SEOs of how they promised to get their penalty lifted or special treatment from an engine because they got a response to an email they sent to a search engineer. Perhaps an even better rule is - don't promise something you can't personally control and deliver.
    _
12. Don't Get Overconfident and Dismiss Other Marketing Channels
    OK, yes - SEO rocks. But don't forget how valuable other marketing activities like email, PPC, CRO, affiliate programs, even display advertising can be for the right scenario. Once you've found the SEO hammer, it's easy to see every problem as a nail - I've certainly been guilty of it. If you can resist, think holistically and provide the best answer from a strategic (rather than tactical) level, you'll become even better and more valuable to your clients.

Your turn - any "don'ts" you'd recommend to fellow SEOs?

p.s. If you haven't read the whole Malcolm Gladwell vs. Chris Anderson with Seth Godin weighing in thing, it's pretty worthwhile :-)