Wednesday, July 23, 2008

Gasta SEM: SXC Marketing

Think you have SEM under control? Think again. Find out how to increase efficiency and lower the boom on the pesky problems of this popular tool.

Your SEM program is terrible. But that's all right because your boss doesn't understand a single thing you do -- he's just concerned that you are doing it and that your numbers look great.

So what are you doing wrong? First of all, stop patting yourself on the back with how great your SEM program is. You're really not that good. Seriously. Now let's move on.

A rock with arms can make SEM look good. Why? Because it's the last stop a consumer makes. By the time they type in a keyword on Google -- and it really is all about Google -- they've already decided what they're looking for. Television, radio, print, banners, events and billboards have all sucked their brains so dry that by the time they come to SEM, they are mindless lemmings.

"I must have product X. Must type in keyword to find product X. Must consume. More, more, more. It will make me happy."

What gets the credit for that last stop? SEM, of course. The consumer sees a television commercial, a print ad, an outdoor billboard and then goes online and types something into Google to find it. Voila!

Did SEM cause that sale or visit? Of course not. It is merely the conduit to it. It's like having an extra door to your store to let more people in. I always marvel at myopic managers who cut their other ad budgets and slowly see their SEM traffic drop and can't figure out why. The trickle may be slow at first, but the curve does become apparent over time. If you want to look like the hero in advertising on the brand side, go into SEM. It's just a shame the creative format requires the writing ability of five-year-old. But it's not about the creative, it's all in the strategy.

In SEM, if you are not maximizing the long tail you might as well be a rock with arms. Why? Google has become the de facto internet navigation engine. Forget portals. If someone wants to go somewhere online, they start at Google. A much smaller entity can compete in the niches of the bigger player's mass reach. It is precisely that "phenomena of choice" that makes the whole AdWords universe work. Unlike in product manufacturing, the price advantages of the long tail are huge in SEM. The fractured niche universe works here, because it's not about someone knowing your brand; it's about them knowing what they want. Going after the long tail in search is different than in product development, chasing after those consumers. As I wrote last week, the costs there can often be quite high, and your business model greatly dictates whether that is advisable.

So how and, more importantly, when do you go after that long tail? Make Google work first
The biggest problem I see with most marketers is diversifying too early. They go on MSN, Yahoo, Google and Ask simultaneously, constantly tweaking each individual program, keyword list and copy. All of that work requires duplicative efforts and drains the time on reporting requirements for your staff. They constantly question why this is working here and not there. It's not strategy, but comparative paralysis.

The Google universe is huge. It is also the biggest market opportunity. Get your entire program to work there first. Build the keyword lists, tweak the copy, expand, optimize and concentrate on making it work on Google. Keep on going down that long tail of keywords. Track performance over months. Expand to the Google Search Network; make the program work there first. Then, once you are hitting the point of diminishing returns, expand the entire program to MSN, Yahoo and Ask. This way it is a strategic approach that is efficient. You just port your program.

Now, are there going to be different performance dynamics across the other engines? Of course there will be. But getting Google to work first is essential. You will end up being able to expand faster than if you work on all platforms simultaneously.

"No time," is what I hear everyone in our industry scream. "Then be more efficient," I scream back. If you can't get your program to work on Google, it's not going to work anywhere else.

Shoot the copywriter, befriend the lawyers
Okay, this is hard for me. I am a copywriter, but seriously, using copywriters for SEM copy is like using a fly fishing pole to catch tuna. It's an efficiency nightmare. A net is coarse, bulky and blunt, but it is a lot more efficient. If you are going after the long tail, where the volume of keywords you will be using can approach the hundreds of thousands, eliminate the process of using a copywriter. A copywriter will grind your process to a halt.

But the copywriter isn't the only drag on efficiency. Remove the approval processes for all copy internally. Your director doesn't need to see it, and neither does your VP. Any approval process on copy just wastes time, a lot of time. And a lot of money, too. It is not about the individual keyword and the copy. It is about the corpus of words. You can A/B test copy within Google all day and keep tweaking it for best performers. A copywriter will never be able to provide copy that is so much better and more responsive that it overcomes the time suck from going after the long tail. The long tail value will far outweigh that. Step back and see the big picture.

In fact, the ones you really want to eliminate from the process are the lawyers. "But legal always has to see the copy!" No one is going to sue you over SEM copy because the search engines will shut you down long before that happens. They are the gatekeepers.

So what should you do with that legal department that is causing you nightmares? Well, befriend them. Sit down with them. Explain the issues, the cost to the company, the time drain on them, the speed to market issues and ask them to draft copy guidelines that you must adhere to. Can't use the word "Guarantee?" Then eliminate it from your copy and post those guidelines in the SEM department. You will save an enormous amount of money, time and frustration.

Look, a legal department once told me that we had to have a disclaimer in our SEM copy. The disclaimer was 120 characters. Ha! Show them the process, the copy and how it works. Set up a meeting with them to walk them through a single keyword posting, copy, the Google Console and then explain that you have 20,000 keywords. They'll get it.

Utilize ad optimization companies like DidIt or Efficient Frontier. You can try and go after the long tail yourself, but the only way you can effectively manage the portfolio as your keyword list expands into the thousands is to bring in some firepower. It's not about getting a click on your SEM ad, it's about getting the conversion on the back end. Remember that the click is irrelevant. What the consumer does when they get to your site is everything. Agencies like Efficient Frontier and DidIt not only have the tools to help you, their optimization technologies can adjust bid pricing for the most efficient conversion. The money you spend with them will never be more than the amount you save by using them. But it does take time for those systems to gain all of the intelligence they need to properly optimize.

I have heard stories of brands saying they tried optimization software but it didn't work. The only cases where I've seen optimization fail have been when companies did not give it enough time, were short-sighted or when companies approached SEM from a test program they did internally that showed better results. The problem with internal test programs is that anyone in SEM can make a small list of keywords extremely efficient. But when it's time to expand that list, the problem becomes unmanageable.

There are only two companies I know who took SEM in-house and ran it more efficiently. Both had unique business models and both designed their own optimization technologies for their business models. Essentially they have their own SEM agency internally. But Bob, Mary and that intern have no hope of making it scale efficiently internally unless they are given the resources to do so, and the one thing they will need most is time. If you are a small shop, agency or brand, you can do it yourself. It will just take a lot more time to scale, that's all.

Okay, so maybe your SEM program isn't awful, and maybe you're not a rock with arms. SEM is all about efficiency. It may not be the ad venue that's glamorous, and you're not going to be able to point to the television or billboard and brag as you tell your friends you did that. But if you start strategically and approach the process diligently, you will see better results with less of a headache.

Sean X Cummings