Every industry eventually goes through the difficult transition from manual labor to automated technology. The good news is that this technology generally results in better product consistently, increased efficiency, and cheaper prices. The bad news is that technology inevitably spurs an arms race that leads to layoffs, margin compression, and product commoditization. Many small companies are replaced with a few large ones. Lots of dissimilar products are replaced by a couple of similar products.
Such will be the path for search engine marketing and survival will depend upon a company’s ability to develop the best technology as quickly as possible. Hundreds of small “mom and pop” SEM shops will either go out of business or get sucked up by the bigger, automated players. This will happen very soon – probably in the next 18 months.
This is certainly cause for concern if you are just starting your career as a budding search engine marketing professional. Will you be needed in two years when massive computers, algorithms, and databases are crunching millions of datapoints to derive the optimal bid for every keyword?
The answer is yes . . . and no. There’s no question that automation will reduce the need for some human staff. Companies will be able to do the same amount of work with 10 people that they previously did with 20. But the key is that there will still be a huge need for talented SEM experts.
When I was in college, there was a lab in the computer science building with a big sign on the door that read “Artificial Intelligence.” The lab right next door created a hand-written sign on their door that said “Natural Intelligence.” Though our ability to create “smart” computers continues to improve, we are still a long ways away from computers that can actually think by themselves.
Anyway you slice it, savvy bid management algorithms are all based on combinations of rules developed by humans. If keyword X has revenue Y during time Z and cost is less than Q, increase bid to M. The computer is an order-follower and nothing more.
Granted, this works very well, particularly when you are dealing with tens of thousands of keywords on dozens of search engines. But computers make mistakes, and – more specifically – programmers who develop search technology make mistakes. As a result, there will always be a need for humans to ‘fact check’ what a computer is doing.
But humans will be needed for more than just routine error-prevention. Humans can infer opportunity where a computer can’t. Here’s an example: the Super Bowl is coming up this Sunday. What should you consider marketing for this event? Let’s see – the Super Bowl is watched by a lot of people – they watch it at home – they have friends over – they prepare food for their friends – the South Beach diet is popular right now – vegetables are low in carbs – it’s currently mushroom season – so you should buy “fresh mushroom” keywords.
I admit, the example above is (more than) a bit silly, but it does indicate a process of deduction that is actually highly complex due to the variability of inputs going into the conclusion. Think about it, in that short paragraph above, my mind processed information about technology (watching TV), cultural habits (with friends), current fads (South Beach Diet), nutritional facts (vegetables are low in carbs), and vegetable seasonality (mushroom season).
Yes, it is true that you could try to build a computer that takes these specific elements into account, but this sort of non-linear thinking – the ability to go down thousands of conditional logic paths – will not be replicated by technology anytime soon.
Thus, for keyword generation, for ad text creation, and for keyword bidding, the human element – the ability to weigh thousands of different, always changing factors instantly – continues to be very important.
Will computers continue to play a greater role in SEM? Absolutely! Will SEM experts be obsolete in the next five years? Not a chance.