The price of sibling rivalry
Your mom needs a new TV. Your brother agrees. But would you bid against him on eBay to get it? Not only would that drive up the price and waste your money, your brother probably wouldn't be happy about it.
As individuals, we see how bidding against a sibling with a shared goal would be counterproductive. But at the corporate level things must get fuzzy, as many large companies bid against themselves in pay-per-click auctions when they have multiple brands, products or divisions to which the same keywords apply.
This internal competitive bidding raises costs to acquire customers for each brand/division and for the company as a whole. These artificially inflated click charges can be avoided through better communication, coordination and bidding strategies.
Problem defined. Search engines generally do not let one company serve multiple ads on a given keyword query. By disallowing this, they create a better user experience. Many of the rules the engines impose on advertisers are in response to overly aggressive bidding tactics they've seen in the past. For example, it wasn't so long ago that different ads for one company would dominate the entire first page of search results for a given keyword. This wasn't a result of companies not knowing about internal competition. It was the product of well-orchestrated bidding (usually by small companies) aimed at gathering the most possible clicks by shutting out real competitors. Given such history, engines have legitimate reasons for prohibiting this type of bidding, and searchers have benefited.
But how does a search engine decide that two bidders represent the same company? The decision is much more subjective than you might think. In my experience, it usually comes down to whether the company can pass the red-face test. If you're one company selling the same product via different Web sites, or you're marketing the same product to different market segments or you're selling different versions of the same product via different sites, then you're probably going to be reprimanded eventually. Bonus idea: If you see competitors locking up multiple positions in your auctions by doing these very things, politely inform the engines. Then sit back and bask in knowing that they'll soon face their own red-face test.
However, if different divisions of your company sell legitimately different products on Web sites that also differ in look and feel, you probably can bid in the same auctions indefinitely. But this is where all the problems begin for large companies. Corporate divisions often have different folks running their online marketing, and those folks rarely talk to one another.
When it comes to online marketing, large companies tend to have difficulty with communication across divisions. Perhaps this is because search marketing is just starting to get the attention of executives. I've yet to see a company with a dedicated team, or even an individual, responsible for coordinating online marketing initiatives across divisions.
As a search engine marketing agency, we at times have been tasked to assess the damage of internal competitive bidding among corporate divisions. An analysis for one company with multiple divisions found its bidding on overlapping keyword sets cost the company well into six figures a year. For that sum you could hire someone whose job is to coordinate bidding across divisions and still have funds left over.
What can be done? The answer lies in communication, testing and strategy.
Communication is critical. As all divisions have a vested interest in saving money by not bidding against one another, it should be easy to get everyone to the table. Then it's a matter of identifying the overlapping keywords. The important thing is not to leave anyone out. If they bid in paid search at all, ensure they're at the table. As paid search is a cost-effective way to acquire customers, smaller divisions often pursue it aggressively. And they are likely bidding on branded keywords because they offer the best ROI. Given that, you can just guess which keywords are also most likely to overlap across divisions.
Testing is the next step. Don't assume that bidding on overlapping terms should be eliminated whenever possible. The key isn't to eliminate the overlap, it's to bid strategically instead of wastefully. Through testing, we have found that multiple ads on the same page can be a waste of money.
But we also have found that multiple ads on a page cause a rise in clicks to both ads that is greater than you would expect from the simple addition of the two. Often, products from different divisions of a company are complementary, so it may make sense to serve a user ads for multiple options. On broad keywords that don't indicate the user's intent, it may make sense to serve multiple ads that offer services at the small business level versus the enterprise level. Multiple ads could make sense in many situations, but you need a strategic reason for serving them and a means of testing, tracking and controlling the bidding.
The last step, not easily accomplished, is developing a strategy that satisfies all divisions. When there is a tested, strategic reason to serve multiple ads on a given keyword, which division gets higher placement? This is where the fur usually starts to fly. The ideal situation for the company is to serve the ad in the highest position that stands the best chance of getting a click that converts into a customer and makes the most profit for the company when they do convert. Not surprisingly, I rarely have seen a large company operate in that ideal scenario, and for obvious reasons as each division wants its ad to be No. 1. More often, the solution adopted is the one found to be the most acceptable by all divisions.
To that end, various bidding strategies can be used based upon your negotiating abilities with your other divisions. The first option is a simple removal of keywords. But as mentioned earlier, don't jump to this option as a knee-jerk reaction; test first. You'll often find that it isn't even possible to make the keyword sets unique.
The second option is time-based keyword sharing. If divisions must not run ads on the same keywords at the same time, they can share based on time slots. These slots could be broken up by day of the week or even hour of the day. While this approach may be difficult logistically, it is effective at eliminating competition. If the products of your divisions attract different audiences that tend to be active at different times, this solution could be elegant. But often a secondary negotiation occurs around which ads run in which slots because if you're after the same general market, different time slots can vary greatly in value.
The third option is to optimize across all keyword sets using one agency or internal unit that has the experience and technology to execute across the board. This is the best option for ease of testing, coordination of efforts and efficiency of optimization. But it also costs the most, in raw dollars and in human talent. If you decide to be the company that optimizes for the good of the whole, this would be the only great option. The dollars saved and the knowledge you could gain about how your customers interact with your products and divisions would be well worth it.