This is a write-up of the eCommerce PPC webinar we recently ran with Google where we focused on how to use predictive analytics, fully optimise your product feed, and structure campaigns for maximum profitability in order to increase sales.
Generating traffic from eCommerce PPC campaigns is easy but maximising sales and profit is far more difficult. Online retailers with large product catalogues have to work especially hard to get the right product listing in front of the right audiences, at the right time and convince them to take action.
In this article, we look at three strategies we use at Vertical Leap to supercharge eCommerce PPC campaigns for our clients. These will help you increase sales, invest your budget where it makes the strongest impact and maximise profitability.
Predictive analytics takes us out of reactive marketing practices and allows us to respond to consumer trends before they happen.
This is important for several reasons. First of all, some people are more likely to buy during certain periods but the real issue is that advertising budgets are finite and, on a platform like Google Ads, your budget runs on a daily basis – so you’re not maximising that budget at the most important periods of the day.
With predictive analytics, we can optimise bids to dedicate more budget to periods when customers are more likely to buy to detract budget during the times they’re unlikely to take action, maximising return when it matters most and protecting your budget conversions are less likely.
For our clients, we pull data from their Google Analytics account and CRM platform to compile a database that includes years’ worth of behavioural data from their customers. From this data alone, you can start to see patterns in purchase habits – for example, the times of the week or day their customers are most likely to complete a purchase.
However, we can take this much further by mapping out this user behaviour and comparing it to external influences. For example, we might start to see trends between rainfall and purchase volumes for specific products or economic trends and broader purchase habits.
With enough data, we can then predict the impact of next week’s weather on sales and adapt bids to maximise profit before a drop of rain even falls.
For a real-world example, one of our clients is an online pet accessory retailer and we noticed fluctuations in dog coat sales that warranted investigation.
Typically, the first thing you look for is the time of day, days of the week and months of the year where sales are highest. For example, you may find people tend to spend early in the morning or during the evening commute when they need some retail therapy. These are important insights but they don’t explain the inconsistent purchase rates for our customer’s dog coats.
By bringing data in from external sources, we were able to find meaningful correlations between external factors and sales volumes for this product. The standout correlation was between rainfall predictions and figures from the Met Office and resulting sales for dog coats, as illustrated below:
We already knew that dog coat sales were typically higher during the autumn and winter months (as expected) but the Met Office data revealed that sales are typically highest in relation to rainfall throughout the year. This is significant in a country like the UK where rainfall is so unpredictable and a whole summer can be a washout.
This data allows us to get granular with campaign optimisation. For example, we saw that dog coat sales are often highest during the days building up to weekends when heavy rain has been forecast a few days previous. We can also see strong correlations based on the weather forecast for bank holiday weekends where people pay particular attention to rainfall predictions.
So we’ve looked at a real-world example of how predictive analytics can reveal major insights for campaign performance. Now, if we apply this at scale, we can see how external factors ranging from unpredictable weather patterns and sporting events to market shifts, economical trends and political events influence sales for our customers.
These are business-changing insights when you shift from reactive marketing, where you’re only learning from past events, to a more pre-emptive and adaptive strategy where you’re using historical data to predict future outcomes – such as how next weekend’s weather is going to affect consumer habits.
With these insights, we can optimise campaigns for events before they happen, taking full advantage of the biggest influences on consumer habits in a way that’s simply not possible with reactive marketing.
The best part is, you can automate most of this, too, so we can continuously pull in data from the Met Office for the next few days and automatically adjust our customers’ bids to increase/decrease based on those predictions.
Some marketing activities are more difficult to automate – for example, creating content for an emerging trend we’ve identified. But, by automating every possible step, we can put all of our resources into creating this content and our predictive analytics system gives us a head start, allowing us to have all of the necessary content in place so it’s ready to capitalise on this trend when consumer habits are at full tilt.
The first concept to understand with Google Shopping is that your eCommerce PPC campaigns are only as good as your feed. You have to think of your products as search ads and adopt this mentality that they’re only going to perform as well as you make them appear in Google Search.
There are six key areas to focus on here:
With these essentials covered, your product feed acts as the basis for high campaign performance in Google Shopping. The data in your feed determines the quality of your product listings and here’s an example of what a well-optimised listing should look like:
Here, you can see we’ve got a quality product image, an optimised title that includes the brand, gender, product type, size and colour, a compelling product description, accurate pricing (including tax and shipping) and a seller rating review score to show online shoppers that this retailer is trustworthy.
All of these elements play a key role in showing your listing for relevant searches, capturing user attention and building confidence.
So this covers the essentials of optimising your feed for individual products but a lot of our customers sell products in groups – either in a bundled purchase or over a succession of multiple purchases.
For example, one of our packaging clients sells a lot of cardboard boxes and customers often buy these boxes along with other products, such as tape and bubble wrap, which presents an opportunity to sell these as bundles through Google Shopping, using the is-bundle attribute.
This way, users who are looking for cardboard boxes can see these bundles with other items of interest included without having to manually search for them.
Another useful way to group products in your feed is by creating product variants with the item-group-id attribute. This shows users product listings that include variants, such as the same product in different colours, as you can see here:
These variants are useful for shoppers who aren’t 100% sure what they’re looking for and could be swayed by a certain colour, pattern or style. They can be crucial for users with specific needs, too, such as shoppers looking for furniture that fits certain dimensions or a new TV with a specific screen size or key features.
In this final section, we’re going to take a look at how you can manage campaign structure in Google Shopping and how different structures impact return on ad spend (ROAS) and profitability.
When we start working with new eCommerce brands that are already running Google Shopping campaigns, we often find they’re structuring their campaigns in the following way:
If we imagine a clothing retailer, this structure involves building eCommerce PPC campaigns around product types, product ranges, brands, gender, product value and other categorical groups. This makes a lot of sense from a campaign management perspective, too, because you can easily run seasonal campaigns or promotions and this structure also makes navigation and reporting easier to work with.
There are plenty of benefits to structuring campaigns in this way and we’ve specified two of them above: managing budgets and predicting ROAS.
However, you also need to be aware of the pitfalls of structuring eCommerce PPC campaigns like this – and the alternatives available to you.
With the campaign structure outlined above, performance results will often show you that some products or ad groups are more popular while others generate a higher profit margin and, in most cases, it’s those high-margin products that sell in the lowest quantity/frequency.
So we start to encounter problems if you’re optimising to maximise ROAS or conversion value. You fall into this trap of achieving excellent campaign performance and all of your reports look great but, in reality, your campaigns aren’t necessarily profitable because you’re losing track of revenue performance.
Here’s a simplified example of what revenue might look like for such a campaign for the first month:
With a set budget of £40,000, you can see that spend on each product varies due to fluctuations in popularity, seasonal factors, stock levels or a whole range of variables. So, while the ROAS and margin of each ad group remain the same, you could end up generating less revenue in month two from the same ad spend.
This variation can occur from one month to the next, generating inconsistent levels of income from the same budget and campaign performance.
The good news is, we can solve this problem and gain control over profitability by using another campaign structure.
The campaign structure we looked at earlier bundled products together, within the same campaign, using ad groups for similar products (brands, genders, product value, etc.). So we can regain control over campaign profitability if we move away from grouping product types and structure them around products based on popularity and margins.
Now, you’re organising campaigns based on profit margins and you can start to optimise for ROAS and conversion value at the campaign level, while retaining control over the profitability of each campaign. So you can assign budget to the campaigns generating the most profit and redistribute ad spend as performance changes while always keeping profitability as the KPI of each shopping campaign.
This allows you to generate consistent profit from each campaign and optimise performance to maintain or increase your returns. The same consistency means you can also predict profitability with great accuracy as you optimise campaigns and redistribute budget – something you can’t do with the same clarity using the first structure we looked at.
So, when you first start using this campaign structure, your reports may look more like this:
Here, you’ve got equal spend across each campaign but you can redistribute ad spend to prioritise campaigns generating the highest profit margins and increase profit for the same total ad spend, as seen below:
In this case, the profit and profit margins have both increased for this month but the ROAS across all campaigns has actually dropped. This shows how we’ve moved away from optimising campaign performance in isolation and placed profitability as the KPI for the collective performance of campaigns in a structured Google Shopping strategy.
If you need any further help with your eCommerce PPC campaigns, contact us on 02392 830281 or email@example.com. Also, check out our guide 8 PPC strategies for retail marketers.
Callum is a PPC Specialist at Vertical Leap.
Categories: PPC, SEO
Let us review your search marketing campaigns and provide a proposal free of charge.