(Matt Cardy/Getty Images)

One size doesn’t necessarily fit all when it comes to e-commerce strategies, writes Conjura CEO and co-founder, Fran Quilty.

Here he details an in-depth five-step plan for boosting your business’s e-commerce offering through data analysis.

In my previous guest column I looked at the rapid growth of e-commerce in the jewellery sector and offered advice on how businesses of all sizes can make the most of changing shopping behaviours.

Advertisement

Success – at least in the short-term – is contingent on using data to understand who your customers are, what they want and how to engage with new and returning customers most effectively.

The growth of digital-savvy brands like Missoma demonstrates that many businesses in the sector are now ready to take their data strategies to the next level to drive competitive advantage.

That next step is data science.

It goes way beyond analysing data on a daily basis to understand what’s already happened; data science allows you to work through what should or will happen through predictive modelling against different scenarios.

This works by identifying behavioural trends over an extended trading period that can be used to inform growth strategies.

It sounds simple on paper, but the reality is more complex and it’s all too easy to become overwhelmed by the sheer amount of data.

So, how can you make it work?

Fran Quilty

1. Ensure you have enough viable historic data

Context is key when it comes to data science. Making accurate predictions is only possible if you have sufficient ‘good’ data to make the connection between inputs with outcomes, a bare minimum equates to at least 12 months-worth.

Good also equates to ‘normal’, so any data harvested from when the pandemic broke out until March 2021 would fall under the heading of ‘anomalous’.

2. Take care over how your data is gathered

Always check the settings on your data gathering tools to avoid coming unstuck. In particular, beware of the configuration of your CRM and stock tracking systems and check they aren’t set to provide only point-in-time data because any old data is overwritten.

3. Make your data comparable 

If you can’t join up your cloud-based and/or on-premise data systems because they are proprietary and/or incompatible formats, then you will never be able to accurately track a customer journey from marketing to point-of-sale.   

Data spaghetti is not insurmountable though, and there are technologies available that can pool the various data sources into a centralised warehouse.

“Making the most of data science means investing in the right tools, but these won’t get you very far without data scientists”

From here, it can be ‘cleaned’ and standardised to provide a single point of truth.

Then a common identifier should be assigned to each customer to offer an accurate view of engagement. This allows you to tailor value-added services and offers – whether that’s insurance, cleaning, customisation, etc. – in line with the likely needs of the individual.

4. Ask the right questions

There’s no place for guesswork when it comes to data science. If you don’t ask the right questions that feed into a specific strategy, then you’ll get nowhere fast.

Start big with business-wide goals that can be narrowed down through measurable questions for different departments – for example, you might start with marketing and scale out to customer services, fulfilment and so-on to get a helicopter view of performance.

The devil is in the detail when it comes to predictive modelling. When you get the questions right, close analysis of historic data trends will uncover the growth opportunities. This sets you up to develop strategies for new products and services, new geographies for expansion, even new business models.

5. Make the right hires

Making the most of data science means investing in the right tools, but these won’t get you very far without data scientists.

It’s a specialism that requires a lot of patience, attention to detail and coding skills. Consequently, it wouldn’t be fair to ask your IT or marketing team to pick it up on top of the day job.

Try as they might, it’s unlikely they’ll ever be able to extract the full value of the data – and that’s an exercise in frustration for all involved.

Data scientists aren’t a magic bullet in themselves though. Unless you have a data culture throughout the business – one that everyone buys into – then internal politics can scupper any data ambitions.

The fact is, not everyone gets data, some may recognise the value it brings but aren’t willing to share and others may not like the spotlight data transparency shines on their department.

“The devil is in the detail when it comes to predictive modelling”

Consequently, the most important data hire is a leader that is willing to take a firm stance and diplomatically force the issue.

Clearly, upgrading your data strategy to factor in data science isn’t necessarily easy, there are a lot of variables that need to fall into place to make it a success.

Unsurprisingly, there are plenty of businesses that end up disillusioned and turn their back on data science. This is typically because they’ve made the leap too early, or they simply haven’t articulated what specific business issues they need it to address.

Early movers can still gain a competitive advantage from reactive data strategies built around daily reporting, but this will only get them so far.

As jewellery e-commerce becomes an increasingly competitive space, it is those able to pinpoint where the future opportunities lie that will be best placed to maintain growth.