This week, eBay released a study into the effectiveness of paid search ads which concludes that, for the auction giant at least, 'the efficacy of SEM is limited at best'.
It's an interesting study, and well worth a read, but the report's conclusion applicable to brands in general, is it just less effective for eBay, or is there something wrong with the brand's PPC strategies?
I asked a number of experts about the report, and whether they agree with its conclusions....
Do you agree that buying PPC keywords has little effect on sales?
Dan Barker, Online Marketing Consultant:
Answering the question 100% literally: no. I've done a decent amount of PPC audits over the last few years and, even where companies aren't managing AdWords particularly well, or are losing money from it - there are usually pockets of it delivering sales.
Even eBay said it was delivering sales. It simply said that once they looked past the standard methods of reporting on ROI, they found it was unprofitable for them.
Answering the question less literally: I've been using AdWords for almost 10 years. It is much harder to make money from it now than it was when I began using it.
Bids have gone up massively over the years, there is far more competition, many of the methods of making it more efficient have been obfuscated (or in some cases removed entirely), and Google has made it easier and easier to 'waste' money.
Optimisation, testing, driving efficiency are all really important for AdWords.
James Gurd, Owner, Digital Juggler:
To ignore paid search, given the available market data, is slightly crazy. A sensible media approach is to ensure part of your search budget is dedicated to paid and then carefully plan your strategy and underpin it with analytics to measure performance, optimise and learn.
With multi-path conversions, you can often find paid search nuggets that don’t look like they’re really adding much value but are actually influencing areas like direct traffic and assisted conversions.
Knowing how paid search keywords contribute to sales really depends on how well you are measuring that contribution and if you’re taking all factors into account. If it’s last click attribution only, it’s much easier to prove that paid search keywords don’t drive incremental sales, though that’s not necessarily the correct conclusion.
See my post I wrote on running a brand test - for one retailer, we found that dropping brand paid search actually decreased generic traffic/conversion from same terms.
Matt Whelan, Head of Search at The Digital House:
Simply put, no. We've recently performed a test for a market leader who dominates the natural results in a particular niche. We switched off their paid search and saw a 40% drop in traffic and, crucially, a 20% drop in online sales.
Whilst not as comprehensive as the eBay study, the data was enough to prove that paid search is very much a worthwhile investment for this client. However, what's true for one advertiser is not necessarily true for another.
Pete Whitmarsh, Head of PPC at Search Laboratory:
Each company has their own story regarding this. There are certainly a LOT of companies out there that make significant profits from their PPC advertising.
We have worked with advertisers who rely on Google AdWords for as much as 90% of their incoming traffic and sales. Without this traffic they would basically cease to exist as a business.
Just because it hasn’t been working for eBay absolutely does not mean that it won’t work for other companies. PPC is very measurable so it is actually fairly easy to tell whether it is working or not for the majority of businesses.
EBay’s business model only allows them to make a small profit from each transaction so, consequentially, they will probably find it hard to compete with retailers who are making a profit directly from a sale.
Sri Sharma, MD of Net Media Planet:
I can’t agree that buying PPC keywords has little effect on sales. From a branded keyword perspective, it has the potential to more traffic than SEO alone and at a higher conversion rate.
One of the reasons is that brands can better optimise their messaging. From a non-branded keyword perspective, PPC can contribute to multichannel sales and new customer acquisition.
Naturally testing and optimisation is needed and what works for one brand may not for another - as can be seen from Ebay’s study. What would be interesting would be to see if there would be any variation in Ebay’s PPC performance across the different search engines Google, Yahoo, Bing.
Paul Mead, Founder, VCCP media:
Of course not. In the same way that I know a few people who don’t watch TV but that doesn’t lead me to question the effectiveness of TV as an advertising medium.
EBay is a big advertiser so the report makes interesting reading but nothing more than that. It has generated more sweeping generalisations about PPC than I’ve read in a long time. The most sensible comment I read in the debate was Google’s official response that every advertiser is different so you should test these things for yourself!
The more interesting question is why a company like eBay would want to publish something like this? Whatever the reasons I doubt that a natural inclination to share their learnings for the greater good of the industry is among them!
Is PPC perhaps less useful for a well-known brand like eBay?
Dan Barker:Alongside the 'well-known' factor, there are many nuances of eBay's business that impact on what they can & can't do via PPC, and how they compare to other businesses:
- It has very loyal visitors and very loyal customers, who visit the site regularly. If you begin participating in a sale/auction on eBay, you naturally revisit the site again and again. That's not true for 99.9% of sites.
- It has the largest range of products on the planet - they sell more individual types of products than any other retailer.
- It's a marketplace, whose sellers do a large chunk of the marketing for them.
- It has enormous 'organic' search coverage, and a large amount around 'high intent' terms.
- It runs an enormous amount of email.
- It makes far less margin on the average PPC sale (or 'newly acquired PPC customer') than (say) a fashion retailer would. Perhaps that limited the types of ads they targeted.
It depends on what eBay is trying to attract traffic for. Perhaps for brand searches using eBay keywords it is less relevant due to its brand recognition. Sometimes big brands turn off paid search, especially when using day parting.
However, I’ve not seen a major retailer permanently turn off all paid search and not once met a digital marketer considering that rather draconian action.
I’d be interested to know how well covered they are for long-tail generic keywords as well as non-eBay brand terms, by that I mean branded products being sold on eBay such as Kingston peripherals in electronics.
Taking the Kingston example, for “kingston 32gb USB” eBay doesn’t appear anywhere on page one in Google UK (competitors like Amazon and Ebuyer trump them) but on some variations of that keyphrase it does have Google Shopping results.
However, not everyone will click on a shopping result. Also, Google Shopping is now a paid search model, so an intelligent bidding strategy would raise its visibility. I’d be amazed if turning off all their paid search didn’t completely screw long-term visits and conversions.
I dont think the size of eBay's brand is actually the relevant factor here, but rather the uniqueness of its product offering.
The report concludes that paid search was of little value for its "most active" customers. For these customers eBay may well be a natural part of the purchase cycle. They will always see if they can find something on eBay whenever they shop.
Ebay's uniqueness has led to these results. Voucher code affiliate sites are often huge brands, and they have become part of the buying cycle in a similar way, but their market is saturated, without too much differentiation.
I doubt they would see the same results as they have competitors who will pick up their lost PPC traffic if they were to switch it off.
Cian Weeresinghe, Director at Nine Diagrams:
Fundamentally the question is around incrementality (does being exposed to a paid ad change your propensity to convert or would they have come direct anyway), which unfortunately attribution cannot answer and only a control/exposed test like this can.
I believe the incremental impact of PPC varies by industry, retailer and the customers awareness of the product, brand and if the brand sells the product.
However, I've done studies which reinforce eBay's conclusion. If the brand is well known and the consumer knows they sell the product they are searching for, the incremental impact on existing customers is limited (but not 0).
Then the question is can you use it purely as a acquisition channel (ie optimise to new customers) which opens the LTV can of worms.
Regarding brand keywords, normally I'd say than the substitution effect between paid and organic isn't 100% as eBay claims and there is incremental revenue to be had bidding on your brand; but again depends on awareness and competitor activity.
In my eyes it's Google's tax on getting SEO traffic for free.
Quite possibly yes, because of two reasons. Firstly, the awareness for this brand is immense with significant advertising across other marketing channels. Secondly the brand acts as a very unique destination, the single largest marketplace where price sensitive consumers can bid to get a great deal.
Is there something eBay could improve with its PPC strategy? It seems to have had a scattergun approach to keyword bidding in the past.
Yes, it probably could. I don't know the ins and outs of the campaigns, and can only see what eBay did from an 'outsider looking in' point of view, but it felt like there were big opportunities missing from the PPC strategy over the last few years.
Here are five things that may have helped them:
- Focus purely on 'new customer' terms.
- Focus on the terms where they have direct competitors.
- Greater segment
- Bid on other people's brands more. John Lewis, House of Fraser, Amazon, ASOS have (without much scrutiny or mention) massively grown their customer bases online using this technique. It works hugely well for well-known 'brand neutral' resellers.
Yet I very rarely saw eBay bidding on other people's brands, or even on '[Brand] [Category]' ads. For example, search google today for 'Ted Baker Shirts' or something similar, you'll very likely see Amazon, House of Fraser, John Lewis all up there. You would very rarely have seen eBay.
- Tying it up with the user journey. eBay know a huge amount about their customers & their visitors, yet the site was always 'one-size fits all' when you landed on it from a PPC ad. Even a blunt instrument like proactively persuading 'non-customers' or 'lapsed customers' to sign up for their emails when landing on the site from a PPC ad may have helped drag up the ROI.
I wonder if eBay had segmented that down further whether it would have concluded there were big profitable pockets of activity that they could have expanded, and big unprofitable pockets that could have been ditched.
Those bizarre keywords were bought by affiliates who were able to direct link. EBay's keyword process was extremely sophisticated and the algorithms and platform beyond anything EF etc had at the time.
However, I'm not sure it's been invested in much in the last few years and it seems like eBay's focus is driving repeat usage from organic traffic via mobile adoption.
It has always been a bit of a bad joke when it comes to PPC. They have regularly been spotted on weird keywords with even weirder adverts.
Clearly, managing a PPC campaign for a company that literally sells everything is an extremely tough challenge and it has used a lot of automation to get keyword coverage as expansive as possible.
However, one has to believe that this could’ve been tackled in a more strategic manner in order to build up more relevance by refining and cutting out areas of the account that aren’t working. I’m certain that if eBay just focused on one category per week it would quickly be able to turn the results.
There could be a huge number of things eBay could do to improve its PPC strategy but its dangerous to speculate from the outside. There are rumours that eBay ‘bulk buy’ or had negotiated special deals with Google for effectively bidding on ‘remnant inventory’. Who knows.
What I do know is that we manage lots of large scale campaigns at VCCP and it requires a different, more technology led approach. If you cover 170m keywords it’s not surprising that it’s possible to find a few areas of what appear to be poor best practice.
However, we can’t use a selection of comedy DKI ads to say that eBays PPC strategy is poor. You could argue that with eBay's automated long tail campaign there is no such thing as a bad keyword or a bad ad – only a bad bid (i.e. one that doesn’t reflect the target ROI for that keyword or campaign).
Ultimately if this study gets everyone talking about measurement, incrementality, attribution and how to do PPC in the right way then that can only be a good thing. But the lesson is find out what works for you and not what works for eBay.