6 common reasons analytics projects fail

 

6 common reasons data analytics projects fail

 

Gartner estimates that around 85% of big data analytics projects fail.  That’s a scary statistic, and not only because it implies that a mere 15% of analytical projects will be successful - it’s also incredibly ambiguous about the definition of ‘failure’, making it difficult for organisations to know what to avoid in order to not end up in that 85%.

I’ve been involved in the analytics industry for over a decade, and though every company and challenge is unique, I’ve seen a number of common themes that tend to trip people up when they’re jumping into analytics.  Here are six of the most common (and significant) reasons why a company doesn’t ‘succeed’ in getting value from their analytics work.


project objectives outlined icon1.  The objectives of the project are unclear

The number one reason for an analytics project failing is because success has not been defined well enough.  Having unclear goals and objectives for analytical work often means that a department or person within the business will end up unhappy with the results, or that they won’t be implemented (this is another issue I’ll go into later). 

It’s incredibly important to spend enough time up front outlining the aims, objectives and problem statement for any analytical work so that all major stakeholders within the business are on the same page.  Matt Wilkins goes into a bit more depth about what this entails in Stage 1 of the Datamine Guide to Data Strategies, which you can download below if you’re interested.

 

New call-to-action

 

data cleanliness icon2.  Organisations jump into analytics without having their data ready

I often see that large analytics projects are built on data that is not well-maintained, whether due to incompleteness, poor cleanliness or a lack of early stage governance.  This often leads to wasted time and resource, because complex projects require clean and usable data to ultimately be insightful and actionable.  Without proper governance in place to guarantee the health and quality of the data, you also run the risk of making business-critical decisions without the best possible information.

It’s important to determine how well-maintained your data is – in doing so, you’ll know whether the insights that are being provided to you are reliable enough to base significant business decisions on.  If some (or all) of the data is not of a good standard, this can continually be improved over time as part of the analytical projects.  We also recommend that monitoring tools are put in place to check for errors and maintain the health of data moving forward.

 

project implementation purple icon3.  There isn’t a solid mode of implementing the project results

An analytics project might fail to gain traction or buy-in from the business before it can move through the initial stages.  This isn’t ideal, but it’s not the worst-case scenario - worst case is when a business spends time and money on a project and ends up not being able to implement or see value from the results.  This leads to unhappy stakeholders, tighter budget and further apprehension towards using analytics.

So how do you avoid this happening? It all comes down to the planning stage. Before you decide whether or not to pursue a given analytics project, make sure you have the value capture outlined. You need to understand what good will look like, why good is important and what action can you take from the project. Taking an agile approach is a great way to do this, as you get to break large projects into sprints and constantly show return to the business. You can view the entire project in stages with a priority focus, while keeping momentum building and stakeholders happy.

 

technology icon4.  The technology is valued more than the solution

With all the buzzwords hitting business these days, it’s easy to get swept up in the idea of implementing some new and exciting technology in your organisation. Don’t get me wrong, I’m not suggesting you shouldn’t be exploring this as an option - just make sure if you do, that it fits the outcome that you’re trying to achieve, rather than the other way around. Rather than saying ‘we have to have this technology’, you should say ‘what technology will help us solve the problem at hand?’

Another common mistake businesses make is over-engineering solutions. It’s easy to get caught up in the intricacies of a cool solution rather than thinking of the output, but it’s important to make sure the solution is straightforward enough for you to be able to get actionable insights. Solutions are only useful if they meet an objective and are built in a way in which they are flexible for change in the future.

 

5.  Over reliance on an internal person or team

Subject matter experts are great, but when you rely too heavily on one person at your business, you run a number of risks:

internal teams icon
  • They’re unintentionally doing things incorrectly, meaning wasted time and resource
  • They leave the company without properly documenting everything they were doing, taking their IP with them
  • They don’t have the particular expertise necessary to get the job done
  • Other internal work takes priority over the task at hand, prolonging the timeframes for delivery

 

When you give analytics projects over to one internal person in the business, they’re often too immersed in their area to bring fresh ideas or different approaches - they can’t see the forest through the trees.  Often business stakeholder needs are not fully understood and the output doesn’t match the true requirements. This is where having an external partner can come in useful. Bringing in a third-party expert who knows what they’re doing, maintains IP and has the expertise to get the specific job done means fewer headaches and less wasted time and money.

 

external analytics partner icon6.  Over reliance on an external partner

With all of that said, relying entirely on a third-party expert for your analytics needs also isn’t the answer. Many data analytics firms try to foster this dependence in clients, but we believe that the sweet spot for true business acceleration is in a combination of external expertise and building internal capabilities. It’s all about choosing the right partner - one who won’t mould your problem to their offering, but rather work with you to identify the right solution.

A number of our clients have large and capable internal teams of their own who handle a lot of the day-to-day analytics work, meaning Datamine gets to come in and support on the new, difficult or complex projects. This enables our clients to build their own teams and IP, yet gain the momentum and expertise of a partner who has been in the industry for over two decades.

 


 

If you’re keen to get started with analytics (or if you’ve already started but you’re ready to start accelerating), Datamine can help. Here’s an article outlining the process of engaging with an analytics partner in case you’re a bit unsure where to begin - we start with a quick chat to figure out whether or not we’re right for you and you’re right for us, so no pressure.

Sound like something you’d be interested in exploring? Get in touch.

 

New call-to-action