COVID-19 has changed how we live — and shop. At the end of March 2020, when the world locked down for the first time, online shopping went up as consumers moved from buying in physical stores to buying online.
Marks & Spencer International — which serves 40+ countries outside of the UK — enjoyed increases in web traffic, largely thanks to M&S’s high visibility for lower-funnel search terms (i.e. branded search queries, like ‘Marks & Spencer’, ‘M&S dresses’, or ‘Marks & Spencer kids’ uniforms’). In fact, branded traffic increased by +85% YOY.
To capitalise on increased online shopping, however, we needed to scale search activity and boost the brand’s visibility in high-funnel (non-branded) searches, too. These included: Generic search terms, Google Shopping ads, and Google Discovery campaigns.
Our solution: we pivoted M&S International’s search budget to make the brand more visible in high-funnel (non-branded) activities.Specifically, we increased the non-branded media budget from 67% to 75% of total spend in 2020 (vs. 2019).
After reallocating the budget, we switched our account/bid management solution from Google AdWords to Google SA360, to access the most advanced bid management tools and bid strategies.
General campaign tactics
Using SA360, we launched:
A Data-Driven Attribution model – to analyse every touchpoint in the customer journey. We ensured that campaign types like Discovery, Generic and Shopping (often underestimated by last-click attribution models) were given more credit for interactions and used machine learning to optimise bids against relevant keywords within our target categories — Clothing, Kids’ clothing, Lingerie, and Homeware.
Auction-Time Bidding (ATB) – Second, we automatically adjusted bids for each auction using ATB, taking into consideration internal and external signals. Internal signals included: Time, device, location and remarketing lists; External signals included: browser used, app used, language, actual query, ad creative, search partners and operating system (iOS or Android).
We monitored search trends in every market, every week, and bid more for when we saw spikes in activity. For example, when we spotted searches for underwear in MENA were up, or activewear in Europe, we bid more for these terms.
Budget Bid and Target Bid Strategies – To optimise budgets across 40+ markets, we grouped campaigns based on similarities in past performance, regions, and promotional calendar. Then we set our level of spend for each week, streamlining the budget pacing management throughout the month. We measured budget bid and target bid strategies against our ROAS target.
Agile ad copy management – Significantly, we created ad copy templates to save time and increase consistency. We pushed weekly events and promotions across 40+ markets, including ads for full price events (including the launch of M&S activewear brand “Goodmove” and celebrity lines, like Rosie Huntington-Whiteley lingerie), and discount promotions (seasonal sales, local events and multibuy offers).
Google Shopping tactics
To boost visibility on Google Shopping searches, we changed our data feed provider to manage a higher number of SKUs (Stock Keeping Units) than we previously had access to.
Using this information, we created feeds and launched Google Shopping Ads in 7 markets for the first time: Bahrain, Saudi Arabia, Oman, Qatar, Kuwait, Italy and Switzerland.
Additionally, we created second-language feeds in a further 9 markets, complementing existing English language feeds. Feeds included native languages in MENA, the Czech Republic, Greece, and French in Canada and Switzerland.
Finally, we accelerated M&S’s transition to Smart Shopping across all markets, so we could automate bidding and ad placement and capitalise on spikes in demand.
By shifting M&S International’s Paid Search budget to focus on high-funnel (non-branded) activities, we boosted traffic to the website by 110% which drove increased website orders
Traffic through to the website from paid activity during the campaign period increased by +110%.
Total orders placed on-site increased translating to significant increases in revenue YOY
Importantly, our high-funnel (non-branded) activity played a significant role in these positive results.
Orders from high-funnel (non-branded) traffic increased 196% YOY.
Conversion rate increased 19% YOY
Share of revenue from high-funnel (non-branded) traffic increased from 9% to 18% YOY.
Specifically, the amount of non-branded revenue generated in 2020 increased +238% YOY.