What Twitter Can Learn From Google (and Groupon)
Over the past few weeks, there’s been much discussion about Groupon and whether its incredible success is coming at the expense of the small businesses that have been running deals through its service. A number of past customers have stated that their promotions were so “successful,” they nearly went out of business. There were three main reasons for this:
- Most were unprepared for the overwhelming number of new customers they’d be dealing with and didn’t have the necessary infrastructure in place to effectively react to them.
- Others had a difficult time turning new customers into repeat customers. Businesses that use Groupon basically discount deals so significantly that unless they can get customers to purchase multiple times, they end up losing money on each one they acquire.
- Finally, some Groupon customers simply didn’t have a handle on the economics of their own businesses, let alone Groupon’s. A significantly discounted product or service — combined with the fact that Groupon takes 50% of the deal price — created such huge losses for some companies that their promotions were pretty much doomed from the start.
The last two points are really the most critical. And yesterday, Nick Saint, in what could be the first SAI article I actually agreed with, explained why these issues don’t mean there’s a problem with Groupon’s model, just that some of its customers aren’t using it properly.
Having founded a search marketing firm in 2003, this all sounded very familiar to me. At that time, search was still very much in its infancy. As we went out to speak with perspective clients, we generally heard one of two things: 1) “Hang on, you mean to tell me there are paid ads on Google?” (which, of course, there were) and 2) “I’ve tried paid search before and it doesn’t work for my business.”
While there was undoubtedly some percentage of companies for which that was truly the case, the overwhelming majority of them weren’t seeing poor results because search couldn’t be effective for them, but because it was a new channel that they didn’t take the time (or have adequate resources) to figure out how to use.
For instance, many advertisers were unaware that search engines provided daily budget caps. As a result, budgets that were intended to be spent evenly over a one-month period, were sometimes spent in a day (or less) instead, with nothing to show for it. Other advertisers were completely oblivious to the art and science of keyword research, spending money on keywords that had tremendous volume, but would never be used by perspective customers. And still other advertisers knew nothing about match types and were therefore taking no measures to control who was actually seeing their ads. The list of reasons why a lack of understanding for the way the channel worked led to poor performance obviously goes on and on from there.
The other issue that led some early search advertisers to the false conclusion that search couldn’t work for them was a lack of understanding around metrics and how success would ultimately be measured. Advertisers who had experience with display were focused on impressions, sometimes setting a goal of always being in the #1 position for all the keywords they were buying. Others asked that we optimize campaigns around click-through rate. The bottomline is that these advertisers took few, if any, measures to improve campaigns while they were running. They just took a look at the results once the campaign was finished, saw that they had acquired few new customers, and declared the channel ineffective.
As a result, we (and many other SEM’s) spent much of the first two years we were in business evangelizing the merits of search marketing to some of the world’s biggest brands and reversing false perceptions of how it worked (or didn’t). At the same time, Google began providing a variety of comprehensive resources for smaller advertisers so they could better understand the benefits of search and the mechanics of how it worked. It was really at this point in time that search began to take off.
I was once again reminded of this experience after Twitter’s recent announcement regarding Promoted Accounts. While I’m pretty bullish on Twitter’s suite of “Promoted” products long-term, I wonder if it may be going down a similar path to the one I’ve described above. To the company’s credit, Twitter has taken a slow, controlled approach to all product launches and has shown no reservations about killing products (e.g. Early Bird) that aren’t performing. Twitter’s also done a seemingly good job of limiting the number of advertisers testing new products and offering the necessary amount of hand-holding to ensure things go smoothly. But the type of scale Twitter’s been talking about creating — and that its investors and the market expect — will require going way above and beyond Fortune 500 advertisers. One need look no further than Ad Age’s recent article about Google’s top advertisers to see why that’s the case — the Top 10 advertiser’s in June accounted for less than 5% of Google’s U.S. revenue during the month. The number of companies advertising on Google is in the millions. It is impossible to hold that many hands, regardless of how large your company may be.
Offering advertisers the ability to purchase qualified clicks — or in Twitter’s case, qualified Followers — can be powerful, but only in the event that potential advertisers first understand how to properly leverage a large following. I’m not sure that the overwhelming majority of them know how to do that at this stage.
It seems clear that the success of Twitter’s Promoted Accounts offering is going to be dependent on scalable resources that help advertisers understand the benefits of Twitter so they can effectively engage newly acquired Followers. That needs to be combined with tools that allow them to properly quantify success. Otherwise, what was previously “I paid for 100,000 clicks to my site and no one bought anything” will be “I paid for 100,000 followers and it had no impact on my business.” And that would be a shame.
Follow me on Twitter: @hershberg