“The first change we had to make was just to make our data of higher quality. We have a lot of data, and sometimes we just weren’t using that data, and we weren’t paying as much attention to its quality as we now need to…” – Ash Gupta, former American Express executive president, Payments and E-Commerce Innovation, LLC
We often hear that data is the new oil. And it truly is a valuable commodity when it comes to making your business as successful as possible. However, an important point to note, and one that is easy to miss, is that not all data is created equal.
If the data you gather is going to be helpful to your company it must be meaningful. That way, you can analyze it and leverage it into successful conversions. The best explanation of conversions is from MarketingSherpa’s glossary, which defines them as the point when recipients of marketing messages perform the desired action.
These desired actions could be clicking on a link in an email, navigating to a landing page and completing a registration form for access to more content, or booking services or buying products. In other words, what you want is to be able to use data to generate conversions of some kind. And you can only do that if the data you have is good.
What makes data good?
By “good”, we simply mean that data is usable and relevant to your specific situation and requirements. So, what makes information a good fit? So-called “quality data” is defined by six key factors: accuracy, completeness, relevancy, validity, timeliness, and consistency. Let’s take a closer look at all of them.
Accuracy
Accurate data describes the conditions it aims to describe. That sounds a little esoteric, but the concept is actually quite simple. Imagine if the data you obtained led you to believe that your customers were mostly males in their 40s, when in reality they were mostly females in their 20s.
This is an extreme example, but it should show you how important accuracy is. Without it, your conversion rates would probably be very low, and your business could waste its entire marketing budget trying to reach the wrong target market.
Completeness
In data quality, completeness means that there are no gaps in the information that you’ve gathered. With incomplete data, it can be difficult to make inferences and gain relevant insights. If participants skip several sections of your surveys, you may not be able to draw meaningful conclusions from their responses. That’s why you should choose the right kind of survey for your target respondents, and provide incentives for completion, such as the digital slots and Scratch Cards Pointerpro offers.
Relevancy
Is the information you’ve collected relevant to your goals and objectives? If not, it’s not good data for you – even if it checks the accuracy, completeness, validity, timeliness and completeness boxes.
Validity
In this situation validity doesn’t refer to the data itself, but to how it was collected. For data to be considered valid it must be of the correct type, fall in the right range, and be in an appropriate format. If your information doesn’t meet these criteria, you might struggle to organize and analyze it.
A good illustration of this would be if you needed to know when visitor traffic to your website was at its highest. You’d need to use software that recorded that information, and would need to decide whether to use 24-hour or 12-hour clocks. If each visit was not noted in the same format, it couldn’t be organized, analyzed or interpreted properly.
Timeliness
How recently did the event that the data represents occur? Are the sales figures the latest ones? Were your surveys completed within the last week or month? The more time that goes by, typically, the less useful and accurate the information becomes. That can lead you to take actions that don’t really respond to the current reality, and thus don’t result in many conversions.
Consistency
Data across multiple sets should be the same in both format and content so that comparisons can be made. Utilising business intelligence software can help organize data and streamline operations, as all data is collected and collated in one central location. This process is made even easier with the empowerment and functionality of embedded business intelligence solutions.
Without the right systems providing a good degree of consistency, different individuals or departments within your company could be operating under different assumptions. At best, this is ineffective. At worst, it can lead to teams inadvertently working against each other.
Using good data to drive conversions
With the quality data that you collect, you can increase your conversion rate in various ways. For instance, if you have a clear idea about who your clients are, you can optimize their customer journeys by creating detailed real-life personas.
Innovative apparel retailer Lululemon Athletic, widely credited with having started the athleisure trend that changed the fashion industry, created personas of their ideal male and female customers. Duke and Ocean were assigned occupations, interests, ages, relationship statuses and even the amount of time they had to work out each day.
Duke and Ocean’s personas influence how Lululemon products are designed and priced, how they interact and communicate with customers, the content on the website, and every other aspect of the company. By using these aspirational figures the brand is able to take customers through a journey from casual visitors to subscribers, to loyal purchasers.
Use surveys to gather good data
How can you be sure that you are gathering high-quality data, and how should you be doing it? Firstly, make sure you design and implement a good collection plan, and set standards to ensure all six requirements of good data are met.
To obtain the new information you could run online or in-person focus groups, ask visitors for input when they leave your website or run surveys. The key here is to think very carefully about what you want to find out, and how best to ask.
Consider whether open-ended questions are best, what images you wish to use, and how you want to phrase the questions. Would you like your respondents to rate what they think of your products, or how often they would use a new item that you’re thinking of adding to your line?
In addition, think about how your data-gathering goals have many metrics you would like to measure. You might want to conduct a cultural or maturity assessment to get a better idea of your audience, or to gauge your target market’s response to a specific issue. Amazon, for example, uses surveys to find out what information customers want to be able to access, and then adds the commonly-requested details to their website to improve the overall retail experience. You should also decide whether to use incentive widgets, and what kind would be best.
Check data-based changes
Split or A/B testing is a powerful tool for evaluating the changes or adjustments that you make to any aspect of your brand. Essentially, you create two groups of customers, one that is exposed to the new situation, and one that is not.
As an illustration of A/B testing, imagine that after looking at the latest customer survey responses, luxury makeup business Sephora is considering a revamp of their landing page for their Gold members. By randomly sending Gold members to the old and new landing pages, and then measuring the conversion rates or purchases within each group over a set time frame, the company will be able to decide which interface to use.
Split testing is superior to simply changing the condition and seeing what happens because it spreads the risk that comes with making any alterations to your website or print media. You’ve also got a real-time comparison between the two conditions, which can be invaluable.
All data are facts, but some facts are more useful than others
To recap, gathering and analyzing data is crucial for any business. In today’s digital age, doing so with online surveys is streamlined and efficient – but only if the data is good and in line with your requirements and objectives. Spend time crafting quality surveys to make sure you get the information you need, and use split testing to evaluate the changes that you make. Above all, review the entire process thoroughly and often.
