44 Foods believe there is a better way to shop.
A way that delivers fresh, seasonal, tasty food direct from the farm to your door. Food that’s packaged sustainably and with minimum waste. Food at a price that’s fair to farmers and families.
So that instead of putting up with poor quality, being frustrated by silly substitutions, facing the nagging worry that prices shouldn’t ever be that low, you can relax in the knowledge that not only is your food the freshest and tastiest it possibly can be but that you are also supporting local British businesses.
They developed a solution so that your food is only sourced after you order it. It’s not coming from a warehouse but direct from the farmers, growers, and fisherman as and when you require it. It is then expertly packaged and couriered directly to your door and can be left in a safe space so that you do not need to wait in.
Alongside our main objective, we also wanted to:
Reduce cost per paid acquisition
Scale paid marketing to increase gross sales
Increase the digital awareness and presence of the brand
Build a foundation upon which to scale more aggressively in peak seasons.
Help increase the lifetime customer value of every new customer
The solution to improve 44 Foods marketing efficiency was a pretty big one. Working across Shopping, Search, Display, and Paid Socials, there was a lot of work to do to get the ball rolling in a few short weeks.
We started where you should always start...with tracking. We went about setting up advanced tracking throughout the site on Facebook and Google Analytics, tracking all macro and micro conversions to get full transparency on what was really happening when a user gets to the site in order to increase conversion rate.
This gave us a huge amount of valuable data which was supremely valuable as the market environment worsened, allowing us to increase the overall site conversion rate by 131% and supplying us key information to make important decisions on factors such as device bid adjustments, landing page changes and time-specific campaigns.
More recently, the rapid uptake in iOS 14 adoption rates in summer 2021 dramatically decreased the amount of data we had available from Paid Socials. We overcame this via integrating more closely with their CRM to allow us to benefit from offline conversion data - improving ad attribution in ads manager by as much as 50%, an excellent outcome that mostly mitigated the long-term impacts of iOS.
To get the social side of things underway, we split up the main product categories, prioritising those that had the highest average order value (AOV) such as food bundles. We built out cold traffic campaigns, supported by retargeting campaigns, both top of funnel (recent engagement with the company’s social media or website) and middle or bottom of funnel (product views or incomplete checkouts).
Once this ecosystem structure was proven to work, we broadened our cold traffic testing to account for other key product areas - Fruit & Veg, Fish, Generic (Homepage) & Meat. This allowed us to present a variety of different product areas to consumers and therefore maximise our appeal to a range of different customer avatars, while testing which types of product drive the cheapest Cost Per Acquisition (CPA) and highest Return On Ad Spend (ROAS).
In each of these we also needed to test what specific creatives, messaging, copy length and call to actions were delivering results inside the individual campaigns to allow us to optimise spend not only on product type, but on the individual ad creatives and audience in each product category so that we could maximise our marketing efficiency and the client’s overall results. To that end, we undertook an extensive amount of testing, with over 250 different creatives, almost 200 copy variants and over 50 audiences tested.
With the growth of cold spend, we took advantage of the increase in traffic to expand the retargeting campaigns, testing whether or not discounts and offers were better at achieving results than messaging without offers that focused purely on the quality, standards and ethics behind the brand.
The best performers from the above campaigns were moved into a CBO campaign to drive greater spend onto proven winner ads and increase the volume of spend and purchases being made in the account.
By the time we had taken over the account in 2021, we had pretty much missed all of the big special occasions to drive traffic for, such as Easter and Mothers Day etc. We know these are big incentives for purchase and can drive lower purchase costs, but we still had BBQ season bundles to prepare for. With the miserable April and May we had in 2021, we developed a regional campaign that was entirely led by the weather forecast in that particular region for the next few days. I.e If the next few days were looking sunny and hot in that region, we would run the campaigns driving purely to our summer bundles. If the forecasts didn’t look great, we wouldn’t run the campaigns. This allowed us to make the most of our BBQ spend whilst we could with people looking to capitalise on any hot weather and bank holidays that were coming up.
Once we started the new campaigns we quickly identified that there were spikes in conversion rate and order volume between Sunday and Tuesday, when the weekend delivery slot cut off. As a result, we built out two separate campaigns, one with “last chance to order” messaging which has proved extremely effective and one that took our best performing ad creatives and pushed them solely during these peak time slots. This allowed us to really maximise our spend and strike when the iron is hot at times that we know people were most likely to order our products.
We rounded out the retargeting ecosystem by introducing post-purchase retargeting. This enabled us to improve returning customer rates by retargeting former customers with our best-performing adverts and rapidly became some of our most profitable campaigns. This also had the added bonus of putting ads in front of our most passionate customers so even if they didn’t purchase, we would accumulate comments such as “just received my delivery, so happy with the quality”. As these ads were implemented via Post IDs, those comments accumulated on adverts that were also running in other campaigns. As such, not only were the campaigns very profitable, but this implicit social proof from customers on the adverts improved outcomes in the cold campaigns.
We found that consumers fell into two pools, either recent purchasers (last 30 days) and long term (60-180 days) and we could tailor our spend and messaging accordingly. Initially, we started off running our best-performing cold adverts to these audiences, but have recently introduced specialised messaging to entice customers back to the brand.
We understand that getting people to change their shopping habits to a more expensive option can be a big task, and doesn’t just happen after seeing a cold Facebook ad. As such, once we had all of the main campaigns live we started filling in the gaps, introducing brand awareness campaigns educating people on why they should buy from 44 Foods. These included:
We also built out a Recipe Challenge lead gen campaign to get people thinking about the ingredients in their food and introducing a multi-phase retargeting system after someone had made a purchase depending on what product they had bought and how long ago they bought it.
All of this allowed us to make an immediate difference to the profitability of campaigns on Facebook and Instagram, before moving the focus to scaling spend through the second half of 2021. Over the course of 2021, ad spend rose from £3,380.81 (January 2021) to £35,034.12 in November 2021, with 825 individual purchases (up from 274 in January).
This was a multi-channel endeavour, with our work starting on Facebook & Instagram and Google Search and Shopping being added to our remit over the year, with Microsoft, Pinterest and Twitter rounding out the platforms as we moved into 2022.
Whilst our social team was working on all of the above, as mentioned our PPC team worked on rebuilding the Google Search campaigns. These had previously been almost entirely smart campaigns, which gave an enormous amount of control and discretion to Google, who may not always act in your best interests! We wanted to regain control of the account, and so moved to a Single Keyword Ad Group (SKAG) strategy that focuses on serving the most relevant ad to the searcher at any time and gives you maximum control over your keyword targeting and ad relevance/performance.
We started by building campaigns for the main brand keywords we could identify to capture anyone searching after seeing an ad, to take advantage of the greatly increased traffic we were producing from Paid Socials. We then split out campaigns into the company’s broad product categories, developed an ad testing strategy across all of the campaigns, and developed retargeting campaigns for each of the keywords.
Immediately upon the creation of these new campaigns, we were finding our search impression share going through the roof and Click Through Rates (CTR) skyrocketing, which was great, suggesting we were getting in front of the right people. We achieved this by focusing on exact matches, refining our campaigns using huge negative lists that were updated daily to exclude any broader, less relevant search terms and splitting out our campaigns into very specific high intent keywords. This allowed us to see a 452% increase in CTR with an average quality score of 9 and reduced the search impressions lost due to rank by 80%, almost eliminating any search lost due to rank and a 10% reduction in AdWords cost per purchase.
Conversion rates were still lower than we expected from such high intent traffic, converting at lower rates than some of our cold Paid Social traffic. This suggested to us that the landing pages we had been provided with needed improving, so we guided the client’s web team in the process of creating specific landing pages for each of the main product categories, to drive conversion rate up even more.
We introduced shopping campaigns at the same time, starting with an all product segmentation campaign via negative keywords, funneling the users into product campaigns, brand campaigns or specific SKU campaigns. While this was delivering a good cost per purchase, we were noticing our average order values were particularly low as people were searching and buying just one or two items. To overcome this we split out the shopping campaigns by product cost, starting with the bundle campaigns and filtering outwards through the beef joints, whole fish, which command the higher price points. We also focused on building negative lists and adapting product titles to funnel specific keywords into the shopping ads. This allowed us to increase the Average Order Value and the ROAS of the shopping campaigns.
From here, we wanted to scale Shopping further as profitability was very strong. We tested Smart campaigns alongside our traditional Shopping campaigns and got very good results, so these were rolled out across the account. From there, we sought to scale by segmenting by product type, but could not match the results we got from a combined approach. We theorised this was because of the dynamic retargeting embedded into Smart campaigns, and the fact that users would be visiting a variety of different product groups. By segmenting the campaign, we were losing out on retargeting when users entered the campaign via product group A, but spent time also on items in product groups B, C and D.
Given this, we stuck with an unsegmented campaign structure and attempted to scale spend and profitability by improving impression share, click through rates and conversion rates by:
This worked extremely well, with the Shopping campaigns providing a new route for 44Foods to be found by wholesale distributors, who would place bulk orders at retail prices, in addition to the steady growth of “regular” consumers. Google Shopping has therefore gone from being an entirely untested platform to representing £6,950 of monthly spend, almost 30,000 monthly clicks and around 200 purchases a month, with a low CPA at £35.
All of the above allowed us to achieve the following results in 2 months: