With a wealth of experience looking after clients’ data as well as The Data Shed’s own, we asked Chief Architect, Stew Norriss, for his advice on how you can harness the data in your business to make better decisions.

Whether you consider yourself to be a numbers person or not, it’s worth taking time to understand the data in your business from a legal compliance perspective and to assist you in making smart business decisions.  

Stew’s advice is to follow this five-step process: 

1) Identify 
2) Quality check 
3) Analyse 
4) Understand 
5) Test 

1) Identify your data sources 

Data comes into your business from a surprising number of places when you stop to think about it. Besides ensuring you are complying with the legal requirements of GDPR and the Data Protection Act (2018), it’s good practice to have an overview of your business’ data so you can spot opportunities and problems early.   

Customer touchpoints for your business might include your social media accounts, your website, contact and payment forms, for example.

If you have third-party tools for customer relationship management (CRM) like Salesforce or Hubspot, email marketing like Mailchimp or Constant Contact, and analytics tools on your website like Google Analytics, then you may use these for data collection and interpretation too. 

2) Check and improve your data quality 

“Data quality is really important. If you have 10 data points in a contact form, you’ll often find that people put in dummy information or leave blanks in at least some of the fields”, says Stew. “The problem with this is if people put rubbish in, you’ll only get rubbish out.” 

This is important to know so that you make informed decisions when you are sending something out physically, so you don’t waste lots of material and postage costs. For email marketing campaigns, the performance data can look skewed if it’s based on poor quality data to start with. 

At a strategic level, before basing decisions on who you should develop new products for, or what customers to target your next marketing campaign towards, you want to be sure that the data on which you base these decisions is high quality. 

Stew advises business owners and marketers to be wary of vanity statistics. “You might think your latest marketing campaign with a data capture incentive has performed really well because 1,000 people submitted a contact record”, says Stew. “But how many of these are duplicates or poor-quality contact records with missing or even worse - misleading dummy information?” 

“These kinds of marketing campaigns can be scammed. It’s not then just the poor quality of the data, it’s also the missed opportunity to have spent that money on something else”, he adds.  

For your website analytics, do a sense-check of how your data is being tagged. For example, is 40% of your traffic really coming from social media advertising or is it mistagged? If something looks more or less than you might expect, investigate it further before making any decisions using this data. 

If you identify that you have a problem with your data quality, stop and take time to improve it. There is no point trying to take important decisions using ‘bad data’.  

3) Analyse data to identify opportunities or pain points 

Once you’ve got an idea of where your data is and feel confident in its quality, you can use it to identify opportunities to fix problems and improve your products and service – both of which can lead to business growth.  

“One of the most important things from a website point of view is where the customer journey breaks”, says Stew. “It’s all about removing friction or ‘pain points’ from the journey. You want to identify where they abandon, e.g. is it payment page, cart, or information pages?” 

You can see this data using an analytics platform like Google Analytics or similar.  

If you have a CRM system you might be able to perform a similar analysis – what percentage of people take the time to make an enquiry but don’t go on to purchase anything from you?  

Clearly, it would be easy to spend all your time analysing the masses of data your business has available and putting lots of third-party tools on your website, or surveying customers at every stage or their interaction with you. But you don’t need to do this, says Stew. In fact, it can lead to a poorer quality experience for your customer, so be careful to only do the minimum you need. 

Stew explains: "The biggest thing people miss is ‘what is the problem I’m trying to solve?’ If it’s that you have lots of enquiries but a poor conversion to sales, start there. Work backwards through the customer journey and check where people might be having trouble buying from you.”  

For example, do they get to a payment page then leave? Start there immediately.  

“Focus on the thing causing your business the most pain. Collect data on that area first and this might drive further questions and narrative. Don’t just try to profile human behaviour; profile problems”, Stew adds.  

4) Ask questions to understand the cause of the problems 

While you’ve been collecting data about a problem, you may have identified a few other areas that might need attention. You can give each a simple rating according to how much ‘pain’ you think it’s causing your business. A poorly-performing payment page would be a high rating. A general information page that people leave would probably be a low rating.  

“Analytical tools can tell you about behaviour, but it doesn’t tell you why it’s happening. If you try to interpret it yourself, you’re likely to introduce bias because you’ve built the system”, says Stew. “One thing I learned at Plusnet very early on in my career and has been reinforced during my time at The Data Shed is the value of getting direct user feedback – getting people in and engaging with the community.” 

Isn’t that expensive, though?  

“It doesn’t have to be”, responds Stew. “We find that people are willing to engage because they want to have their voices heard, so it can actually be pretty easy to get people to come for feedback sessions, just by providing a lunch or some beers, or simply by asking.” 

Stew advises that you use your data to pinpoint the problem and set expectations for outcomes from the workshop session, to stop them from becoming a ‘free-for-all'.  

“When we’ve done this before we’ve had people say about the area we thought was the problem, ‘this bit is really good, leave that alone; it’s this other bit here that’s the problem because...’ so you get some really great insight from the sessions and it saves you a lot of guesswork.” 

5) Test solutions 

Once you understand the problem and the reasons for it, you may also have some helpful suggestions for solutions to fix the problem from your focus group.  

If more than one solution was suggested, again you might like to rate them. This time consider the cost and ease of the changes suggested and assign a rating to each. You can then start with the cheapest and easiest to implement and work your way up until the problem is solved. 

“A/B test the solutions by sending people down different journeys, but don’t make it 50/50 as then if the ‘solution’ isn’t good, half your visitors are getting a poorer experience. Instead, make it 90/10”, advises Stew. “There are many tools I have used previously such as Maximiser or Frosmo that can help with doing this for website changes, and you might need a developer’s help to make the changes, which should be factored into your costs.

“It’s really important to keep people engaged after your workshops and update them with the developments that their input has helped achieve, so they feel a sense of having made a difference”, Stew concludes.  

Once you have fixed your biggest problem you can work your way through the others by repeating steps three to five. 

Stew Norriss is Chief Architect at The Data Shed, based in Yorkshire. The company specialises in helping clients with a wide range of data-centric projects to realise their most aggressive objectives in a timely and affordable way.

Neina Sheldon
Article by Neina Sheldon
Share Article