As consumers, we just accept it, don’t we? ‘£10 Minimum Card Spend’. But why? Well, because the retailer gets charged by the card issuer – about 40p per transaction for a small shop to be more precise. But why? That’s four percent of your tenner! What costs does that cover? Why do they change so much?
The answer, of course, is ‘because they can’. Visa and Mastercard have a 96% market share between them. They don’t even need to conspire together: as long as both of them play it safe and don’t cut their fees, they can go on making obscene profits and the consumers and retailers will just keep on paying.
Google Ads isn’t like that, though.
Google has an 88% market share all by itself. If you don’t like the price that Google is selling clicks at, well… there’s always Bing! That’s why we’ve always been skeptical about automation in general – letting Google be both the buyer and the seller of clicks is a little like handing your wallet to the shopkeeper and saying “could you choose what I need?”.
They won’t bleed you dry, but you’re unlikely to get great value for money.
And, as automation rolls out, that’s indeed what we’ve been seeing. Here at Boom, many of our e-commerce clients’ costs per click were up 20-30% year-on-year this Black Friday, and more broadly, some CPCs more than doubled in the last four years. And the final nail in the coffin: we’ve yet to see a Smart Shopping campaign which reliably outperforms the manually managed equivalent. Maybe, as Google claims, automated campaigns do perform better than manual campaigns on average, but we’ve found that with good data analysis, we can still beat the machine!
So, We Shouldn’t Use Google Smart Shopping?
Au contraire, Pierre!
Google knows more than you. Google is more reactive than you.
Smart Shopping uses machine learning to look at all the signals available (device, location, audience lists, etc.) in real time. There are only so many hours in a day, and the laborious task of splitting out product groups and setting bids and bid modifiers often isn’t worth it any more. There’s still work to do. You can’t just set up Smart Shopping and forget about it: budgets and bid strategies, product title optimisation and feed management still need doing, but realistically, embracing automation will cut your workload substantially. Think of it like a dishwasher.
Deciding whether or not to use Smart Shopping is all about balancing competing needs. For some campaigns, in some accounts, the time saved by handing the keys to Google outweighs the money saved from having better control over which clicks you purchase.
When Can You Outsmart Smart Shopping? And When is it Better To Continue With Standard Shopping Campaigns?
When the market you’re in is changeable.
Google suggests that Smart Shopping campaigns will typically take fifteen days to pass through the learning stage, and another two to three weeks to optimise. It often takes longer. If you’re in a fast-changing market, if there’s a lot of seasonality, if the client’s priorities are constantly shifting, then you’ll constantly be throwing the system back into learning mode, so you’ll be better off managing bids manually.
When just a few of your products account for the bulk of sales.
An oft-quoted axiom in business is that ‘80% of sales come from 20% of products’. Reality isn’t quite so neat: some clients have a handful of star products which account for the bulk of their sales, whereas for others the revenue stream is spread more evenly. If your particular account does indeed obey the Pareto principal, it makes no sense to manually manage all your bids. Take out the top sellers and place them into manual campaigns, with the remaining products going into Smart Shopping.
When reporting and accountability are important.
Smart Shopping currently keeps even the most basic information about where your money is being spent tightly under wraps – no search terms, no audience data, no channel data, no nothing. And no control. If your client or your boss wants to know why we haven’t spent more on Product A, or why we’ve increased spending on Product B, what would you tell them?
This product has been performing really well. We get £50 back for every quid we spend. We’re spending less on it than we used to. Why? No idea! What can we do about it? Nothing!
This product has never sold anything. Traffic gets an unexpected boost the week before Black Friday, and hasn’t gone down much since. Why? No idea! What can we do about it? Nothing!
When you want to push specific products.
As this excellent study from Crealytics confirms, the single most important factor in the Shopping ad rank is probably product price. If you’re the cheapest seller of a particular item, you don’t need to bid high to get your ad shown widely. So we need to be proactive. If we can find out in advance which products we’re going to be price competitive on, we can manually put bids and budgets into place ahead of time instead of following the data. Likewise, if a client has excess stock which they simply need to shift quickly, these products will need to be placed in a regular Shopping campaign.
When some queries are worth more than others.
One of the most powerful tools we have for driving quality traffic to our product pages is query sculpting: using multiple campaigns and campaign priority settings to bid more for higher quality search terms. For example, one of our clients sells gift cards which can be used at a range of popular restaurants. Brand searches like ‘Nando’s gift card’ convert at nearly twice the rate of more generic searches like ‘restaurant gift card’, which in turn outperform the broadest queries (‘gift vouchers’). Using manual Shopping allows us to bid more for better quality searches, and to block entirely irrelevant ones (‘Primark vouchers’). If you need to focus your spend in this way, for example with less mass-market, more niche products, Smart Shopping is best avoided.
When you don’t have lots of data.
Smart Shopping has a minimum requirement of at least twenty conversions in the last 45 days, but works better with far more. If you sell just a few dozen products a month, it’s better to stick with manual bidding.
When some customers are more important than others.
Smart Shopping gets half-decent results by bidding up on people who are more likely to convert. So anyone who’s a “basket abandoner” , or has converted in the past, or has simply spend time on your site before will be bid up. If you’d prefer to spend more of your click budget acquiring new customers, that can’t be done with Smart Shopping.
Should I Use Smart Shopping?
Sometimes it’s better to pay for things by card. Sometimes cash is more convenient. There is no binary choice with automation, either. Because generally our clients want to minimise their costs, and Google wants to maximise its revenue, there’s a fundamental gap which good account managers can exploit. There’ll always be opportunities to use manual bidding to improve profitability. We just need to spot them.