From both a business and analyst point of view, being overloaded by data is inevitable. Endless amounts of data is collected on a daily basis, particularly in an increasingly digital world where consumer data can be captured at greater speeds and volume than ever before. It’s impossible to avoid data overload entirely. However, where you can avoid overload is by being smart with how you use it. With that in mind, I’ve come up with 5 top tips for the analyst to escape data overload.
Before you even start thinking about analysing data, it’s important to have one clear objective in mind. Ask ‘what is it that I want to find out?’ and ensure this is well defined and measurable. This could be a matter of asking yourself this question when confronted with the data, or making sure the person you’re analysing the data on behalf of has set a clear target beforehand. What you want is one clear goal and one hypothesised outcome to test.
| Tweet This | |
| "The first step to avoiding data overload is defining a clear objective." |
When wording the objective, avoid the use of 'and', e.g. “Do customers purchase X and when?” This disguises two objectives as one. Instead, keep it simple and have one top-level objective, e.g. “Do customers purchase X?” Stick to this target throughout.
Now that you’ve come up with an objective, you can start to think about what information will be the most helpful in answering it. In particular, which data source is of more importance than others and can offer you the most insight? Consider your objective and sieve your data. Keep only the relevant data and remove the rest. Even when left with the relevant data, prioritise the best source. Ask yourself ‘which will be the most useful in answering my objective?’ and stick to it.
For example, if your objective is, “How much do customers spend on their summer holiday?” you’re likely to choose transactional data over online browsing habits or use of branded tweets. Transactional data tells you the exact quantity of what they spent. The other two, however, can only inform you of their purchase intentions or brand opinion. Whilst these may be interesting on another day, on this occasion keep it simple and focus on one source.
When analysing the data, don’t get distracted. Set a time limit and stick to it. If you’ve not been given a deadline for the analysis, give yourself one. If you have, bring the one you’ve been set to an earlier date. Not spending too much time on the data will prevent you from losing focus and investigating data that doesn’t inform your objective. Encourage yourself to make quick decisions as this will help simplify the process. No doubt you will have a lot of good information at your disposal but having a deadline means you only consider the best.
After analysing the data, your findings will most likely be overwhelming in size too. It’s your responsibility to simplify this information, making it clear and presentable to stakeholders. New research by Esrl UK conducted on 1000 adults across the UK showed that 60% found maps and graphics easier to understand than text. Therefore, the use of graphs, tables and charts is clearly one way forward.
| Tweet This Stat | |
| "60% of UK adults find charts & graphs easier to understand than text." |
Of course, this also depends on the client’s preferences as well as the topic in hand. A financial director, for example, may prefer numerical data whilst a marketing director may favour more visual diagrams. As for the topic, if your objective is more numerical such as “How much do customers spend on X?” it will naturally lend itself more to tables and charts. A more open question, “Why do customers buy X?” will require more detail and explanation in response. Before reporting the results, ask your audience how much detail it is that they want and avoid including too much if unnecessary. Find out what their preferences are. The form of presentation they find easiest to interpret will likely help them digest data better and make a more informed decision.
After all of the prioritising, analysis and presenting of the data, you might actually find that you don’t have enough of the right data to answer your original objective. The data you’ve been given may be valuable from another perspective; telling you, for example, when customers purchase a certain product, how much they purchase and how often. But you may still lack understanding of customers’ underlying opinions or drives. Despite the volume of the data you have, it could be that you need more to help you understand why customers are behaving the way they do.
These are only a few tips for avoiding data overload, I’m sure there are plenty others. Can you think of any other approaches? Do you currently use any of these already when analysing or presenting data? Let us know in the comments below.