Using data science to help SMEs adapt to the new normal

There are a huge number of unanswered questions. Has customer behaviour radically changed? What social distancing measures will remain in place? Will supply chains function normally? Do the same marketing tactics work? How can we prepare for a prolonged economic turndown or, worse, a second wave? What does all this mean for the way businesses should lead, engage and up-skill their people?

The list goes on and on. Many SMEs will of course just be happy to get their doors open and start trading again. However, top of mind will be how to remove some of these unknowns quickly. This is where data science can play a critical role.

The vast majority of SMEs can afford basic data science expertise.

First, let’s broach the common misconception that data science is only for bigger, wealthier companies. This is of course incorrect. The vast majority of SMEs can afford basic data science expertise and they are likely to benefit substantially from the insights they gain. There is a wealth of research showing the clear ROI of data science for nearly every business. It is therefore critical that SMEs consider data science, or at least data analysis, as an essential function.

At its core, data science helps businesses make better decisions by telling them not only what is happening in their company and the wider industry, but also what is likely to happen. It is the predictive quality of data science that is particularly valuable right now.

Of the questions I listed above, the first one regarding consumer behaviour is the most fundamental. Businesses need to know as quickly as possible if their customers - or prospective customers - have altered their buying or marketing preferences. The answer to this question will tell them how they need to adapt their company to survive and even prosper in a post-lockdown world.

Right now, there is a real interest from consumers to share data

So how do we tackle this question? Let’s begin by recognising that the current situation has revealed just how weak many companies’ customer or client segmentation is. Now is the time for SMEs to engage with their customers and improve it proactively. This can be done through zero party data collection, ie, data that customers proactively give you.

Right now, there is a real interest from consumers to share data to enable them to get a more personalised service and sitting at home they have more time to engage. This data capture will enable you to be much more focused and personalised in the coming months, which will lead to better results.

The second priority should be understanding what has happened during the lockdown. For companies that were entirely shut down this could mean gathering any scraps of information that are available, for example, website visits, social media interactions, calls to contact centres or even engagement on any email marketing campaigns that continued to run. For businesses that could continue to operate online, the job is easier and it centres around identifying the actions of current and new customers, such as what did they buy, when did they buy it and how much did they spend?

When this information is gathered SMEs can begin to understand the current situation. For direct to customer businesses, my recommendation would be to start by undertaking a new churn analysis. Initially, this will tell an SME which customers remained loyal, which are at risk and which they have likely lost.

This information can be contextualised by running another churn analysis using data from the same period the year before. By combining these results, we can begin to see the segment of customers that are likely to have deferred their purchases because of the pandemic.

Survival will depend on maintaining the pre-lockdown customer base.

This data will immediately tell an SME who to target with a re-engagement marketing campaign. Survival in the first few months will depend on maintaining the pre-lockdown customer base. Data science again plays a very important role here. Further analysis of the churn results to reveal the profiles of the individuals within these groups will dictate the marketing message and approach.

The analysis may reveal particular groups, perhaps less tech-savvy individuals, or those from industries that would indicate they were furloughed or made redundant, that may have deferred purchases. Tailored messages can then be created to re-engage them, your improved customer segmentation will support with both the messaging and when it is appropriate to contact them.

For a post-Covid re-engagement campaign, SMEs will be working without a net. They will not know what deals they should offer, if they should provide incentives such as specialist shopping hours for at-risk groups, or the best way to reassure different segments about their healthcare measures.

This is where another great quality of data science can come to the fore - the philosophy of testing and refinement. The various messages and deals can be tested with their respective segments, the results analysed and the most effective strategy quickly identified.

Further experiments will also reveal if an SME’s customer communication preferences have changed since the lockdown. This can range from determining when people now read marketing messages and what channel they prefer to deeper questions around whether their customer journeys have changed - but more on that later.

Data science can be used to identify potential future value.

Of course, all of this is only taking into account online behaviour. The real pressing question for SMEs with physical presences is how will customers act in store? Gathering data will be critical. In an ideal world, data that identifies customers could be collected at the point of sale via card linked loyalty programmes. However, for most SMEs this won’t be applicable.

The next best way may be using more traditional methods of surveying both in store and online customers. Data scientists can set the parameters of these questions and then analyse the results to create a broad picture of how these customers are behaving. This data can then be fed into the wider marketing and commercial strategy. For SMEs that have captured new clients during the lockdown, data science can also be used to identify their potential future value, further helping to refine these strategies.

Many questions will also be raised about how to operate in a post-lockdown environment. For many SMEs, the biggest risk will be a significant portion of their team getting ill or having to self isolate at once. Data science can help predict the different scenarios here and reduce the risk.

Business intelligence can reveal where potential risks are in the supply chain.

From a logistics perspective, an SME with multiple outlets may wish to understand which outlet they should open first or what level of service they can and should offer. Business intelligence can reveal where potential risks are in the supply chain. Using predictive modelling, SMEs can be warned when they are likely to be pinch points and operational research techniques can support logistics planning to optimise supply chains.

As the weeks progress after lockdown, it will be easier to determine whether behaviour has temporarily or permanently changed. A clear indicator of this could be revealed by assessing if and how customer journeys have altered.

A marketing attribution model which compares the route to purchases for different customer clusters before the lockdown and on an ongoing basis after it was lifted may be the most effective tool. This business intelligence will then enable more long-term commercial planning.

Third-party data sets can also be used to take into account the impact of wider trends. For example, sector-specific retail sales or footfall figures will help a shop owner ascertain if any downward trend or even temporary spike in sales is unique to their business or part of the economic impact of the Pandemic.

The ideas I’ve outlined are of course only a few of many different ways SMEs could use data science to adapt and thrive. The most important message, in my opinion, is that in a manifestly uncertain world the clarity and insight data science provides is incredibly powerful.

Right now, the Government is modelling different scenarios for the UK and businesses need to do the same - data science is the only way to achieve this with any confidence. For data scientists, the next few months and years could mark the moment where the sector becomes truly mainstream, not because it is a fashionable added value service, but because it is absolutely fundamental to the survival of businesses.

Natalie Cramp is CEO of data science company Profusion