Part 1 of “Finding New Customers” addressed why new customers are important to a brand or retailer.
Part 2 will cover how to find new customers from among the choices of online media programs that are in widespread use. Note that all of these channels are effective at yielding sales, but some drive more new customers than others. Here is a rundown:
Comparison Shopping Engines (CSEs): Consistently one of the top media sources for yielding new customers – often in the 80%+ range. CSEs are somewhat on the decline, but they can still represent up to 5% of online sales. The datafeed work needed to participate in the CSEs can be unwieldy, but there are some good, inexpensive sources that can not only maintain the feed on an ongoing basis, but also create it, if needed. The best feed management companies also optimize the feeds for all of the CSEs (and PLAs), so picking a feed partner whose capabilities match to a retailer’s needs is important.
Non-Brand: Broad non-brand keywords generate the highest proportion of new customers. Many searchers initiate a product search on a broad non-brand term and subsequently search and convert on a trademark term. Unless a brand or retailer can stitch the click of the non-brand term to a sale on trademark, it will appear in tracking as if the non-brand keyword did not yield a sale. It is a Catch -22 in that shutting down core assist non-brand keywords for not hitting target ROI goals will ultimately limit sales on trademark terms. There are many ways to tie the tracking together, but the industry has been slow in adopting it.
For those who actively run broad non-brand keywords, some choose to loosen the target ROI goal because they find it beneficial to pay a premium to acquire a new customer. There is a bit of analysis needed to tie new customer percentages to the actual non-brand keywords but it is very doable and definitely sensible if new customers are a priority. Paying a premium for cart starters, which has assist value of another type, is also a good tactic.
PLAs: These ad units are also a good source of new customers. The conversion rate typically exceeds that of non-brand text since the clicks deep link to the related product page on the site, which is a requirement by Google to run these ads. (The downside of this is lower average order value which is usually higher if a searcher lands on a category page that displays other merchandise beyond what the searcher was looking for.) Note that a similar dynamic occurs as that of non-brand text ads in that many searchers who first view a PLA go on to convert on trademark. Like non-brand search, analytic solutions can tie these two together.
Trademark: This is the most efficient segment of paid search and, surprisingly, the new customer percentages often look better than one would think considering how many people use trademark terms for navigational purposes. The reason is that with last-click tracking, trademark search will get the credit for sales that likely originated in other marketing channels such as TV, display, print, OOH, non-brand search and PLAs. Without these other channels driving sales down the funnel to trademark search, the new customer percentages would look much lower.
Display: The percentage of new customers from remarketing, which consists of prior site visitors, is generally low. When a marketer runs prospecting display, which presumably is all acquisition, it builds up the cookie pool which ultimately gets remarketed to, so that accounts for some of the lift in the new customer percentages one would see on remarketing. Most marketers use some form of view-based tracking when measuring display, so all but the click-based orders (usually a tiny percentages of the sales) get double counted in another channel. Because of this, it is hard to get a read on the specific value of display on new customer generation; although, many are striving to figure out a methodology that works for them to track it (and great solutions do exist).
Social Display: Twitter does not, as of yet, generate acceptable ROI for direct marketers, so most of the marketing opportunity with social right now is with Facebook. Facebook performance keeps improving as they continually refine their ad offering. That said, the best performing ad units – typically the promoted post, custom audiences, and remarketing – all attract a high proportion of existing customers rather than new.
If I had to choose the top online marketing channel for its propensity to yield new customers, it would definitely be non-brand search. The fact that someone has searched for a product indicates they are already interested. Good ad copy, a relevant landing page, competitive price and offer are the hooks to grab the sale. And if a brand or retailer is not visible on a non-brand search, the sale will likely go to a competitor, which is not optimum. If low performance is the reason a marketer is not participating in non-brand search, it is well worth the investment to implement a different tracking solution that can tie the search on one term to a conversion on another. This is the key ingredient to making non-brand search perform to its full potential.