April 23, 2013: A recent study released by eBay, “Consumer Heterogeneity and Paid Search Effectiveness: A Large Scale Field Experiment” refutes the value of paid search and claims an overall ineffectiveness both of branded and non-branded keyword ads.
Over a specified test period, eBay conducted a series of SEM campaign experiments that stopped running paid queries for brand based keyword strings in search results to determine if organic results picked up the clicks of the forgone advertisements. In addition, eBay selected a group of test cities in the US to cease all non-branded keyword advertisements as well, their tests initially found that in these areas revenue was not affected by the cessation of paid ads.
- Branded visits were almost entirely “replaced” by organic results.
- Non-Branded keywords had a negligible impact on sales and revenue.
If we were to take this study as canon, we would all be seriously questioning our ad budgets. However, each case and each advertiser has its own unique picture and properties, and I’d like to refute some of the things in this study that eBay takes for granted.
According to eBay’s study, when a company is not advertising for branded search terms, click traffic is completely replaced by organic search traffic. This is probably true. Someone going to the level of searching a specific brand name is well aware of the company and fully intends to visit the site. Despite this reality, there are still several benefits to advertising for your own brand, especially for B2B marketers, including:
- The High CTR% associated with branded search can have a positive impact on account wide quality scores.
- This can allow you to pay lower CPCs on other keywords, as well as potentially achieve higher ad positions.
- In B2B leads generally need to be nurtured much more than in the B2C space.
- A branded search may be one of these many touches, and showing up in paid results can reinforce trust.
- Remarketing (Google or other) can tag paid users differently, to help with the nurturing process.
eBay is unique in that the site itself acts as a search engine. I disagree with some of eBay’s branded search choices. Terms like “eBay shoes” don’t necessarily make sense or provide a good experience to a search user; someone looking for “shoes” is better off just searching on eBay’s site and browsing directly through the different categories.
eBay may not find any value in bidding on its own brand, however it could be valuable for a smaller or niche auction site that is trying to take market share from the giant. Bidding for competitor brands is more important for smaller or less glamorous brands that searches wouldn’t for explicitly by brand name.
In the B2B space, when decision makers are in RFP phase, they often research many vendors. People searching for a competitor is potentially a ‘warmer’ lead and may be further along the sales process than other potential customers. In certain sectors, it could take several months for a lead to become a prospect or have a dollar value assigned to it, anything that can be done to mitigate some of this and reach a true ROI quicker should be examined.
eBay has over 170 million keyword strings, several of which that were showcased in the study were exceedingly broad and vague. Keywords like “cellphone” and “memory” are weak and poorly targeted; these types of searches are almost better to be made on the eBay search platform than on a search engine. This is just poor keyword selection which could certainly be a major factor in EBay’s lackluster results.
On top of poorly relevant keyword, eBay is probably overpaying for them. The quality score associated with these searches is undoubtedly low due to the broadness of the term and the unrelated page copy, which negatively impacts page rank and cost. eBay’s item turnover makes it difficult to have highly relevant landing pages and keyword selection. It’s hard to replace cellphone with “Samsung Galaxy S3” (a much more targeted term) because the landing page is always being filled with different products.
For some lesser sought products inventory may even disappear from time to time. Due to the sheer size of eBay’s keyword list, there is probably numerous inefficiencies, poorly targeted keywords and ads, and just overall ‘waste’ that hurts their bottom line.
Finally, I question eBay’s definition of how revenue and sales were created from the paid channel. eBay reportedly measured sales that happened “within 24 hours” of a paid click. It’s an auction site and often sales can take days, perhaps a more relevant success metric should have been chosen such as:
- Acquisition metrics (accounts created)
- Interest metrics (time on site, pages views, items views, on site searches made, bids)
- Social metrics (clicks through to social channels, etc.)
What About B2B?
In B2B it’s quality over quantity; less is more. A PPC campaign should have very limited use of broad match keywords, few phrase matches, highly relevant and ROI focused keyword lists, and ad creative should be crafted to verify only the best and qualified traffic comes through.
The findings of a recent study from Optify further supports the use of PPC in B2B marketing campaigns due to its considerable contribution to site visits and leads generated by search ads.
When your customer base is so small and fits such a unique and hard to fill profile, it’s imperative to maximize impressions and clicks from potential users. While large brand names like “eBay” may not reap any major benefits from branded paid search queries, smaller brands still do. A “win at all costs” advertising strategy is used in new customer acquisition and this can’t be achieved without making your branded keyword relevant, targeted and visible in search results.
Images and graphs courtesy of eBay study