We all know we should be collecting data, the question is what data should you be collecting and how can you use it effectively to grow your business? We’ve gone from having no information to having so much that we don’t know what to do with it. We lack the skills to manipulate and interpret it to form predications that validate our intuitions.
Most of us think of it as a technological issue but it has more far reaching consequences including improving operations and customer relationships, creating better products and processes and delivering competitive intelligence.
• Identifying data relevant to your business
• Sources we can collect data from – including online and offline
• How to organise our data to identify correlations and causality
• Scoring data & KPIs – how to identify metrics that matter
• Data management tools
Special thanks to our event partner LobsterPot Solutions… improving your data story.
Why should organisations care about data?
There has been much research into organisations that make decisions based on data and those that don’t. The results indicate that organisations that do use data are four times as likely to introduce products to new markets and three times more likely to increase their growth significantly.
The biggest disruption today is not technology, it’s people - citizens, customers, employees. We have incredibly high demands. When you enter a shop you expect the retailer to know who you are and to pre-empt your need before you realise you need it in the first place. You might have only 10 seconds to capture someone’s attention before they walk away and if you don’t know who they are and what they want before they come, you may miss that opportunity. That all comes back to data analysis and decision point.
You don’t need huge amounts of data – too much and you can get lost in ‘organising’ it. We have the benefit now of around 30 years of experience in what companies have done wrong so we can learn from those mistakes and not waste time on things that aren’t important.
What can we learn from these 30 years of experience in data collection?
Historically people often started with data first and tried to find what insights could be gained, but by the time they identified information of value, the business had moved on or the insight was not relevant anymore.
Today we start with the business outcome first. What is the question we want to answer? How do we define success? What is the KPI that we need to measure? Once we understand the question, we can work back to determine what data we need to capture. We then choose the technology and decide where should the data come from.
How has data processing changed?
We’ve seen a big shift in how data is processed. Around 20 years ago businesses had to write code for data analytics, then it transitioned to ‘drag and drop’ and now we are moving to the next iteration - cognitive. This is where you load a spreadsheet into your data tool and then ask it a business question like, how has my business grown year to year?
Instead of learning how to use the cognitive tool we are essentially teaching analytical tools how to understand our language which changes how you interact making it more pervasive. Both IBM and Microsoft are currently developing in this space.
What tools are available for processing data?
You can process and analyse a phenomenal amount of data in Microsoft Excel. This is an ideal entry point as it comes standard with most computers so there’s no real cost and it’s incredibly powerful.
If you require more power, then look to invest in an analytical tool.
IBM offers a great free cognitive tool called Watson Analytics where you upload a spreadsheet and it analyses it for you.
What should I consider before choosing a data processing tool?
Before you procced with any investment it’s important to ask:
- What do I need to know?
- Do I have information that I can get that out of?
- Could I get this information with the tools I already have?
- Will I actually use it?
In a nutshell it’s about assessing how you will benefit.
What are the limitations of data?
Ultimately it’s a human who has to deal with the data and if they don’t know how to read it or can’t understand it, then they won’t use it. This is particularly an issue for customer-facing staff or those ‘delivering’ in businesses who haven’t been taught how to read a simple bar graph.
If you don’t make the data simple enough to understand or overwhelm them with a huge volume, then they won’t have the confidence to make decisions based on it. Communication and education is key.
There is almost an assumption or an expectation that people know what to do with data but usually they don’t. People are often making decisions based on KPI’s in a hierarchical structure. We need to skill up individuals so that they know what actions they need to perform.
You may often also find that people, particularly those that have years of experience in a role, will be hesitant to trust data. The more they are exposed to the data, the more opportunity they have to realise it’s power and will begin to rely on it.
How important is design in data?
Often people don’t know how to present data in a format that is easily understandable and if it can’t be quickly interpreted within 20 seconds then they are likely to consider it useless and disregard it.
If it doesn’t look pretty then people won’t get it – they’ll either suspect that you’re trying to hide something or it’s going to waste their time as they have to work to get the information.
That being said, the data has to be there. Taking a bad report and making it look good doesn’t make it a good report.
There also needs to be a balance in design between looking good and giving the user what they need. Five years ago everyone started using flipping charts because they looked sexy. The reality is that users would only flip it once and then become frustrated because they couldn’t get the results instantly.
Ideally present users with a one-page summary which clearly depicts the story, allowing people to interpret and conclude what you are conveying.
How can we overcome differences in data interpretation when modelling?
As data models become more complex, training of the model is critical, as is the feedback loop.
The difficulty comes when different stakeholders want to interpret things in their own way as it’s challenging to inject that human interaction onto a model. This is why the focus for major data processing players is on cognitive solutions - it’s about examining how I can effectively communicate with my computer to have it tell me the information I want, and then how the system can learn and automatically add this back into the model.
There’s an element of both top down and bottom up feedback - so as an organisation you have to agree how you describe certain metrics. For example, how do you define a customer - is it someone who has made a purchase or is it someone who walked into the store? You need to have that common understanding so when you sit around a table and talk about business performance you are all talking the same language. Organisations need to have standards for defining metrics, they need to ensure they’re collecting quality data and to provide stakeholders with the flexibility to view it the way they find easiest to comprehend.
Where can a small business source market research data?
There is a huge amount of public data sources available as well as subscription based data sources. You may also have to be more abstract in your search. Data is not static and its changing all the time.
If you’re paying for the data make sure you perform a cost benefit analysis to ensure that this particular data and what you’ll get from it, is really worth paying for. One of the best free sources available is social media as it allows you to listen to what people are saying in the market. This might include running surveys or looking at what’s important to customers of competitive products and services to identify trends. Not every industry can do this, but for those that are B2C it’s much easier. If using social media make sure you’re using the channel where your target audience are - Facebook, Instagram, Snapchat and Twitter all have different demographics).
What tools, techniques are processes can enable informed decision making?
In terms of motivating a human, the most powerful motivator is reinforcement – hence we need reinforcement for every action we carry out. This comes from a feedback loop. For our organisations to become agile, which is required in order to keep up with the speed of the market, we have to relook at how communication flows through them as currently it’s too slow.
This means we need to empower people from the bottom up, particularly as these are the people interacting with customers on a daily basis. As the eyes and ears, they are the source that can confirm information received from data collected. Effectively utilising them to confirm data will allow businesses to finesse their processes. Ideally this means having a process where data comes both down and up through a business.
Cascading data at each hierarchical level of an organisation also helps – it’s about giving the right person the right amount of information to make their decision and the person above or below that the appropriate amount of information.
It’s key to understand your stakeholder’s requirements and how they interpret data. You need to engage by creating a data story that addresses their needs rather than your own.
You also need to be consistent in the way the data is presented. Without this it makes it difficult to compare data.
How useful is collecting data from location tracking, say from Pokemon Go?
There’s two parts to this:
- What you can technically do with the data; and
- What you can ethically do with the data
Technically, if a business knows where you are then they could send you an offer to come in to their store which you tweeted about last week.
Ethically speaking, many companies would not send location based information unless the customer has explicitly accepted to it because they don’t want to creep you out.
What impact is data having on jobs, industries and wider society?
We are now creating technology that can read and understand millions of pieces of content or data and produce evidence based recommendations that provide a percentage based degree of confidence in the recommendation along with a list of the supporting evidence.
The implication is that the job market is going to change. Research jobs – whether it be medical, legal or other, will become less important compared to jobs where people are actually interfacing with other people.
Each decade as technology evolves, it impacts the jobs market. In the immediate future, cognitive and artificial intelligence are going to be a massive driver of change. What we need to be cultivating in schools is a passion for learning as moving forward people are going to have to do that for the rest of their lives. Instead of teaching kids how to code, teach them how to learn, how to use technology, how to create, how to innovate and not get too tied down into a particular skill set.
It’s important to remember that data and technology are tools for us – to serve us and not the other way around. At the moment we haven’t got the balance quiet right, certainly from an organisational perspective. There’s a trend towards Social Default Bias where rather than make decisions based on our own analysis, we are significantly influenced by the decisions other’s make. As we become more technological the brain becomes lazier because it doesn’t have the energy to make decisions.
As our brains become overloaded we gives away consciousness and cognition. More of our decisions are made in auto-pilot mode. A side-effect is that business decisions have slowed down considerably. For example, the sales process for solutions is extending out to 18 months because businesses are selling to committees - where every man and his dog needs to have a say but nobody makes a decision because they don’t want to take responsibility. We are at an interesting juncture where we have all this nicely presented, analysed data however we’re failing to be productive.
Another example of how data is changing the way we live is the Internet of Things (IoT). Devices installed around homes are capitalising on the fact that we are tired and don’t want to make decisions. Learned behavior means we can be prompted by delivery food services with our favourite choices – what we order every week.
The IoT is using data to improve how we behave and consume. For example, devices can tell us the cheapest time to heat our homes which allow us to save and alter our impact on the environment.
What is a marketing cloud?
A Marketing Cloud is a suite of products that allows you to make personalised offers in real time and is automated. It can allow you to do everything from managing campaigns to optimising promotions and understand pricing lessons.
It allows you to define business rules and to introduce data including historical transactions, customer data, social media and to record other interactions. It also has a large variety of third party data’s and incredible connections which can help you design these interactions.
How are changes in data processing impacting government and being used during election campaigns?
It’s safe to assume that data plays a huge role in how parties analyse consequences of proposed policies and preference flows. Social media especially allows parties to gain a good indication of support for policies. The feedback loop is instant. No longer is mainstream media deciding how policies are interpreted. It means parties can continually shift and tweak until they reach some level of traction with the voting public.