To say 2021 has been tumultuous for retail would be an understatement.
Brexit, HGV driver and CO2 shortages, China’s factory slowdown, skyrocketing shipping costs and price increases have pushed many companies to the very edge. With labour shortages rife across the UK, even when stock has eventually arrived, getting it out of ports and into warehouses has been nothing short of a logistical nightmare.
These challenges have affected retailer’s bottom lines and left lasting reputational damage irrespective of size. Retail giant John Lewis found four-fifths of its TrustPilot reviews were rated bad or very bad resulting in it coming under fire for reneging on its 92-year-old “Never Knowingly Undersold” price promise leading to a rather vague value promise replacement this year. [1]
What is clear is the retail status quo has irrevocably changed and competition has never been fiercer. The first half of 2021 saw a 49% increase in advertising spend YoY, as companies deployed retained profit straight back into customer acquisition. Knowing this is a decisive moment to sweep up competitors, companies have done all they can to be in prime position to outdo one another on price, promotion and offer.[2] This year’s high-street Black Friday lacklustre performance offers even greater evidence that in 2022 all roads lead to e-commerce and with that, marketers have a serious challenge on their hands to cut through a crowded online marketplace [3].
Marketing is now a 1st Party Animal
Despite the doom and gloom, throughout the pandemic, retailers have enjoyed a flood of new online customers and by osmosis, have captured a wealth of customer and behaviour data.
In the year ahead, marketers will be challenged to drive ever more efficient and effective campaigns. Sheer weight of impressions and brand awareness can only be bought at unsustainable prices. To avoid an unwinnable battle for audience attention, Data analysis is the smart road around the problem. By analysing customer data and pulling those insights into campaign targeting and messaging, big wins at the tactical level of CTR and CPC bubble up into overall ROAS gains for the data-savvy marketer, while everyone else just doubles their bids or deepens their discounts.
Data science is at the core of this effort.
Using state-of-the-art techniques that go far beyond usual marketing metrics, data science enables marketers to get to the very root of answering why a customer buys from them, who they are, how to best target them and where new valuable customers can be found.
The days of data science being available to the privileged few are over. Data tools enable companies to go toe-to-toe with the Goliaths of retail. With huge developments in cloud computing and visualisation, truly data-driven marketing is well within reach for ecommerce managers with modest budgets.
Truly data-driven marketing
From data enrichment to natural-language processing to postcode analysis to name but a few, there is an abundance of opportunity to move away from traditional and sometimes clumsy marketing methods and engage consumers using razor-sharp targeting and messaging to secure the greatest ROI and in the process gather even more insight.
But what are these data science techniques?
Data Enrichment
Data enrichment is a method of considering a single data source and overlaying it with other complementary data sources to help build a more in-depth picture.
For example, an online food retailer that wanted to dive deeper into a customer cohort that had come to their attention because they were characterised as buying in bulk and consistently. Using the purchase history in customer data they could see this cohort exclusively bought 12 packs of 1.5 litre bottles of sparkling water. They lived across the UK and used various payment methods none of which lent itself to any further conclusions that this cohort were purely fizzy water fanatics.
At this point, data enrichment comes into its own in helping get a clearer picture. A data scientist would start applying other data sources to try and get a clearer picture of what characteristics aside from fizzy water this cohort has in common.
It could be as simple as, when overlaid with topography data, that all these customers live up steep hills and don’t want to lug their fizzy water bottles up them. Or you could find that by calculating the average driving distance between the cohort’s home addresses and local fizzy water merchants that there is not one in easy reach.
With this type of insight, you could then build an automation that invited any customer who lived up a steep hill and outside of a 20-minute driving range of a large supermarket into a VIP Fizzy Water programme. You could also build an ad campaign with specific messaging, a bigger budget and target it at people who live in postcodes that are on steep hills and outside easy driving distance of a fizzy water merchant to find new customers who you know are likely to behave similarly to your fizzy water fans.
At Vuzo, we dive deep into your data to uncover insight that goes beyond the arbitrary metrics of AOV or LTV to gain an in-depth understanding of customer behaviour that you can easily turn into actionable insight.
Natural Language Processing
Actions speak louder than words, so the saying goes, but at Vuzo we think words can offer companies incredible insight into the minds of their customers. Companies spend huge amounts of time and resource on their tone of voice and brand message, but what if these could be data-driven?
By ingesting all communications your customers have with your business, we can pinpoint the language your customers use to describe your company, products, and service.
Your current tone could be factual and informative, when in fact your customers use far more emotive language when talking about your products.
From what product features customers really care about to listing a colour as turquoise as opposed to teal, our Natural Language Processing analysis will enable you to ensure you are emphasising the things that are top of mind for your customers.
On average, our clients have seen a 65% increase in CTR and 20% increase in conversion rate with language identified in our Natural Language Processing analysis.
Postcode Analysis
Location data has long been heralded as useful by marketers. Yet, solely relying on Google and Facebook’s suggested geographic targeting and their subsequent ‘black boxes’ leaves retailers at a disadvantage to be able to confidently build tests they know they can rely on and accurately measure performance.
The locations of your customers understood beyond just revenue or frequency of purchase gives the ability to be laser-focused when it comes to targeting. Considering customer behaviour with regards to the type of products purchased, customer lifetime, retention and attrition rate enables you to identify geographic areas of interest which you can use to your advantage when it comes to allocating marketing budget.
This kind of targeting means you can focus your marketing efforts for maximum return on investment as well as give you the ability to build marketing tests you can trust.
For example, if a company found itself with a surplus of high price point beds, you can leverage postcode analysis to identify the areas in which you know you have a high volume of existing customers who based on previous purchase behaviour are likely to be amenable to seeing high-end bed adverts. This type of insight means you can also reach new customers which based on their geographic location you know will behave similarly to your existing customers.
Our GeoLogic tool goes a step further and enables our clients to use their existing customer postcode data to identify areas that share the same characteristics (type of house, average income etc) as their traditionally high performing postcodes and identify entirely new and untapped areas where do not have customers, safeguarding ROAS and avoiding spend wastage.
This insight goes further than just digital advertising. Our clients have used postcode analysis for direct mail campaigns as well as identified the best OOH advertising opportunities. In some cases, it has even helped underpin location planning for opening new stores.
We cannot solve problems with the same thinking we used to create them, Albert Einstein
Post pandemic, the retail landscape has been forever changed and so too should retailers’ approach to its customers. A company’s greatest asset is its customer data to preserve and crucially increase its market share. Marketers must look for innovative ways to use this data to expand and deepen their knowledge of their customer base.
At Vuzo, we have worked closely with our retail clients to develop data-science tools that help them answer difficult questions and help them build and create truly data-driven marketing strategies that have a real impact.
Looking to 2022, there will no doubt be new challenges to overcome but by leveraging the power of data science techniques to underpin your marketing, retaining, and finding new customers will not be one of them.
[1] https://www.ft.com/content/72cac67e-caaa-4391-ada9-58c1dcd717c5