[nickelcitymario]

Pedantic correction: "Ogilvy".

While the only metric that matters is sales (I fully agree), you need other metrics to understand how those sales happened. This is the dangerous thing.

I've worked at companies where people were evaluated based on very specific metrics, like web traffic or social media engagement. The problem with that is you end up focusing on improving those numbers rather than improving sales.

But if you only focus on sales, you can't tell which activities made a difference (or in which direction). You need other data to understand what's happening.

A better book than Ogilvy on Advertising, in this regards, is Claude C. Hopkins' Scientific Advertising. It lays out all the principles for testing ads, back when we didn't have any concept of analytics. The basic method was: Try different things in different markets to see what works best. Keep doing it indefinitely, so that your ads only ever get better.

There's no reason the same approach can't be taken with public relations. If you're working for a national brand, try assigning some PR folks to exclusively work the media in one given region. Then see what the difference is.

(Obviously, you have to do a lot of work to "control the variables", but any data scientist -- or any scientist at all, actually -- should be able to figure that out in their sleep.)

It's not rocket science. These methods have been around for a very long time.

Honestly, I think we've just gotten lazy. Setting up market tests is a lot of work, whereas creating an A/B test in AdWords is easy. But there's no such thing as an unmeasurable publicity tactic (whether it's advertising or public relations). You just gotta be willing to do the work.

[AlbertCory]

Great message.

A famous quote from an executive (not one in advertising):

"I know half my advertising budget is wasted. I'm just not sure which half."

[nickelcitymario]

John Wanamaker said it, I believe. Both Ogilvy and Hopkins quoted him as well :-)

[sosborn]

With PR, the effects of a failed test can take down a company (or, the existing c-suite at least). That's why people tend to stick to a "tried and true" playbook.

[nickelcitymario]

That's true of any test of any kind... Don't try stupid things.

"Tried and true" means doing things the same way everyone else is doing it, i.e. following the crowd. The whole point of marketing/advertising/publicity/PR/other-words-for-the-same-thing is to stand out from the crowd.

So while I agree that people tend to stick to the "tried and true", that's also why most of it doesn't work. You can't both stand out from the crowd and follow that same crowd. They're contradictions and mutually exclusive.

But that doesn't mean you should lose your head and do stupid things just for the sake of doing something different. Small changes, implemented incrementally, are the best ways to enact long-lasting change. You have to go into every test fully prepared for it to fail. If failure would be catastrophic, that's a stupid test.

[sosborn]

> You can't both stand out from the crowd and follow that same crowd.

I agree - but PR is about so much more than just standing out from the crowd. If you are talking about PR in the marketing sense - sure, but things like crisis management, investor relations and internal communications are different ball games.

[AlbertCory]

That's true, but those things apply mainly when the company's attained a certain level of success.

I was in two startups, and for both of those, 99% of the PR person's job was to get the company favorable press coverage.