5 common data project challenges and how to avoid them

Any data project is filled with challenges; solving these problems or avoiding them altogether is the key to success. Narrowing this list down to just five was a challenge in itself!

Challenge 1: My data is getting too complicated

Our mantra is ‘keep things simple’. That applies to the amount of code we write as well as reducing unnecessary maintenance and management of technology. Unfortunately this isn’t everyone’s experience. For example, when you’re attempting to integrate a variety of existing systems it presents your average developer with a tempting challenge: ‘I know how to fix this,’ they immediately shout out, ‘but it’s going to take a while!’ Chances are, somebody, somewhere has already solved that exact same problem. Rather than waste time and money reinventing the wheel you can simplify these things by using out of the box connectors and automated data pipelines which allow you to connect to your source systems straightaway. They’re pre-built and remotely managed, so you can be working on your data in minutes not months! 

Challenge 2: Nobody’s using my dashboard

The dashboard is the place to splash your important findings in a way that is accessible and easily understood. So how come nobody is looking at it? It’s a gripe we often hear from stakeholders who’ve asked for the dashboard and the developers who sweat over them. The truth is, you can make the best, most functional dashboard in the world but nobody will look if it’s carrying the wrong information. The solution lies not in greater technical or creative prowess, but during the planning stages of your entire data project. First define the overall strategy and goals for the business and then ask how your data can help reach those goals. Secondly, identify those individuals in the business who are accountable for the numbers and targets. Without clear ownership of numbers and without insights that actually drive action, then why should anyone care what’s on the dashboard?

Challenge 3: The tech team is arguing with the business team

Traditionally there’s been ‘them’, the business team and ‘us’, the IT guys, and communication between the two can be more dysfunctional than a bad tempered debate in the House of Commons. Cross purposes and hidden agendas can make a long term data project a fraught affair, but there is a straightforward solution: nominate one developer with the appropriate people and communications skills to be embedded in the business to act as a single point of contact. This ensures that information and requests get passed on to all the relevant parties in a language they understand and you’ll avoid tedious delays and costly repetition of work.

Challenge 4: My data project is going over budget

This comes back to our golden rule of keeping things simple. Many data firms will spend three months or more setting up infrastructure and installing software before they even begin to analyse data. You needn’t do any of that if you use a data streaming service like Fivetran or a cloud data warehouse like Snowflake. Rather than wait until day 91 of a project you can start integrating data straightaway And the quicker you begin loading and looking at the data, the sooner you’ll start to see results. Beyond this we like to use the agile approach in any project, delivering in regular stages which gives us the chance to make mistakes and reset if need be. As we like to say: fail fast, but fail cheap.

Challenge 5: My data project is taking too long

This is perhaps the number one gripe for many companies involved in complex data projects. The causes are many, and can be solved by successfully overcoming the other challenges on this list. Making sure you clearly understand the requirements up front, and have a clear view of the problem you are trying to solve is key. In addition, we recommend clear planning with measurable outcomes at every stage and employing a good project manager to keep everything (and everyone) on track and to deadline. Find out how our own agile approach to analytics could help you quickly get the data your business needs. 

Evolution not Revolution – An Agile Approach to Analytics

A commonly quoted adage in technology consulting is that “the last 20% of the work takes 80% of the time”. A report that will be used by a wide audience will often be designed by one or two key individuals that represent the wider business. However, due to this responsibility there is often the impression that the report must be fully completed with all functionality before it can be unleashed to the wider business. This can result in it taking a much longer time than expected to deliver new reports. However, by delivering the ‘must have’ information quickly and enhancing the report over time there can be significant benefits realised.

Faster Time to Value

By analysing the entire scope of the requirements given by the key business users it is possible to determine the core functionality which represents the Minimum Viable Product (MVP). If you are able to develop the core 80% of the functionality required within the initial 20% of the overall time that it takes for the full scope to be delivered, then there is a huge amount of benefit that the business users are able to get by starting to use this core functionality. This ensures that users are able to start getting value from the reporting product much quicker than if they have to wait for the full scope to be delivered, with the “nice to have” requirements often representing a significant proportion of the total development time.

Increased User Adoption

Another common challenge in analytics projects is user adoption. Too frequently there is a lot of time creating a bunch of reports which nobody actually uses. Users can often be frustrated with the amount of time that it takes to deliver reports, and the danger is that if this takes too long they will find alternative solutions to their reporting problems. However, using an agile approach can massively help with increasing the levels of user adoption by delivering the core reporting functionality to the users as early as possible. Additionally, the MVP will be simpler than the final report, meaning that it is easier for users to start understanding how to use the report. This also means they will be more likely to use the more complex functionality that might be delivered in a later iteration.

Better Quality

Lastly, and most importantly, the overall quality of the final reporting deliverable is likely to be much higher using an agile and iterative approach. The fact that business representatives often give the initial requirements can also mean that some key requirements from the wider business community are not captured. It can also be difficult for users to visualise and understand the reporting requirement without actually getting their hands on the data and starting to interact with it. However, by socialising the MVP version of the report you are able to receive much more feedback at an early stage of the development from a much wider user base. It can also be a lot easier to adapt to change and reduce the amount of rework that is needed. For example, if you’re baking a cake it’s a lot easier to adapt the ingredients when you’re mixing the cake than it is if you’ve added all of the ingredients and put the cake into the oven. This is very much the case when developing a report as often the functionality will build upon everything that has already been built, meaning it’s harder to change at the very end.

Using an agile approach to building reports can often lead to a solution that both meets the needs of the business users better and delivers value quicker. Viva la evolution!

How will IR35 impact your data project – and what can you do about it?

There are many things to consider when running a major data project. Weighing up the long term benefits over sometimes considerable short term expenditure can cause project managers sleepless nights. But come April 2020, many businesses will be asking themselves whether they can even afford to keep their project going. 

The latest factor to consider is IR35, which might sound like a secret branch of the security services, but is in fact a piece of HMRC legislation that will affect many thousands of companies and 230,000 freelance contractors. It aims to deal with the problem of ‘deemed employment’, the practice of using workers on a self-employed basis, often through an intermediary company when permanent employment would be more appropriate. Using these ‘disguised employees’ can save companies considerable sums in tax and National Insurance contributions, and deny workers of employment rights. Under the new arrangements it is up to the employer to assess whether anyone working for them falls within the bracket of IR35. 

On paper this sounds like a fair way of tackling tax dodging and unscrupulous employment practices. But in the public sector where IR35 has already been rolled out many IT projects have been put on hold due to fears over rising costs and investigation by HMRC. Understandably the legislation has caused considerable trepidation across the business world. A survey by Be Digital UK found that four out of ten businesses are considering phasing out contractors altogether. The result of this could see many IT projects grind to a halt.  

What can you do to lessen the impact on your data project? 

The rules surrounding who does and doesn’t fall inside IR35 are rather opaque. There is no one definitive rule, rather a number of questions that you need to answer to assess a contractor’s status. Many employers remain concerned they may still fall foul of the legislation. The most foolproof way of ensuring you avoid the IR35 trap is to make everybody employees. Of course this is a significant long term investment, and could leave you with the additional problem of what to do with them once a project has come to an end. 

Terms of employment 

A key thing to consider is whether the role a contractor fulfils falls firmly within the parameters of the project they have been employed to do. It is often common practice for people to shift around as projects change. Contractors can find themselves on the organisation chart next to regular staff and being moved to different parts of the business. This is definitely not OK under the new rules. 

Milestones 

Project based work, particularly in IT is commonly based on milestones, rather than days worked. Projects billed on milestones are great for ensuring that contractors are clearly delineated as contract workers, rather than slipping into the area of ‘deemed employment’. 

Think big 

Peace of mind can come from working with a large consultancy who can guarantee they only use their own people; this circumvents the IR35 headache for the client. This has been the go to solution for many public sector organisations, but it’s one that comes with a considerable price tag. 

Think small 

You could reach out to smaller IT consultancies, as under the new legislation businesses employing fewer than 50 people and turning over less than £10.2 million annually will be exempt. If they need to hire contractors the rules won’t apply to them. Furthermore, they are likely to offer significantly lower costs than the big consultancy firms and can still deliver great value, especially for specialised pieces of work. 

Undoubtedly the introduction of IR35 will cause anxiety, particularly in the short term, about the viability of certain projects. There is no single best way to deal with the new IR35 legislation and ultimately the right choice for your data project will depend on a variety of different factors, however if you would like to speak to Snap Analytics to see how we can help then please get in touch: 

[email protected]

5 things to do before starting a data project

You’re about to start a big data project. Fantastic! We’re big believers in the fact that every business can gain a real competitive advantage through analysing their data. It’s why we do what we do. 

But just before you go running off all excited, stop for a moment. If you really want your data project to be a success, you need to think about five key things before you even start. 

Understand the problem that you are trying to solve 

Chances are you’re looking to data analytics to fix a specific need, something which is causing inefficiencies and costing you money. Don’t assume though that by shaking the data tree enough times a solution will magically fall into your lap. First you need to look at your existing systems to see what exactly needs fixing. It’s only once you have a clearly defined vision and end point in mind, that we can see exactly how we can help. 

Define what success will look like  

Having identified what you want, it is time to think about what a successful outcome might look like. It helps nobody to embark on a data project without setting any specific goals or measurable outcomes. So make a plan, draw up a list of milestones, devise ways of measuring what’s happening and then track the results against that. One useful approach we’ve found is to run a user survey six months down the line and find out how people are using, or benefitting from the findings.  

Align with company strategy 

It’s all very well you dreaming up fantastic, innovative data driven projects that will change the very fabric of your business and the world generally. But it might be best, first of all, to check that your goals are something that fall in line the wider strategic direction of the business. Is this a problem you should even be solving? Is it a business priority? Will it help tick some important boxes when the annual report comes round? If the answer is yes to all the above, fantastic – you’re on your way to getting managerial buy-in and tapping up a healthy budget for an important piece of work. 

Data for the people 

You’ve addressed the needs of the bigger cheeses but don’t forget about the little guys, the people on the front line who are working hard to produce this data in the first place. Think about how this is going to benefit them in the long term, how will it make their day-to-work work easier, more efficient, or more effective? This is particularly pertinent if your business is going through a restructuring process. We’re great believers in the power of data analysis, but if you’re losing half your team it might not be perceived as the best use of company resources. 

Build the right team 

One commonly held assumption we come across is the idea that data is purely a tech led process: you identify a problem or need and the nerds crunch the numbers. It’s not that simple of course. To produce an effective outcome, you need quality input from people on the business side, members of the team who can provide insights into how the company works and what its goals and strategies are. You should bring together people who use the data in different ways and can provide the broadest possible range of experience. That way the insights we produce will be deeper, richer and ultimately more valuable. 

How to get buy-in for your data project

Despite the talk of a data driven revolution, the reality for many companies often lags someway behind the ideal of a business built on reliable detailed information analysed using AI. According to a Mckinsey Digital report on leadership and analytics, CEOs cite their biggest challenges to investing in data are “uncertainty over which actions should be taken” and “lack of financial resources.” Fundamentally, it appears that some business leaders still don’t believe that analytics have a high enough ROI.  

If you know that a data project could deliver huge value for your business but you’re struggling to get anyone else to appreciate the importance of all that expensive tech and numbers nonsense, here are a few ways to help change their mind. 

Speak their language 

There’s no point trying to convert anyone into data evangelists at this stage, instead work within the parameters of your organisation. Align your project to existing business priorities and show that you understand the CEO’s strategies. The best way to do this is to create a link between the results of your project and the financial benefits. Get those graphs at the ready! 

Remember that this is a business transaction and not a technical pitch. You’re wasting your time if you can’t convincingly demonstrate that you are addressing a particular business need, so resist using too many technical terms. Instead, explain using visuals which demonstrate that the outcomes from your data project align with the strategic objectives of the business and will bring tangible benefit to the company and the individuals who work there. 

You may need to convince people that you are addressing a demand that they didn’t even know they had. By the end you want not just buy in, but for them to believe that it was their idea all along! 

Recruit key players 

It’s not enough to expect the techies to wave a magic wand and sprinkle stardust over the business. Successful data projects are a partnership between the project team and key business users who work with the numbers on a daily basis. Don’t expect that a diktat from the 17th floor is going to be enough to drive them into making this thing a success, it’s your job to involve as many key players as possible, make them feel they’re being listened to and that they have some control over the direction of the project. 

Throughout the course of any data project, key business users should have frequent and regular opportunities to provide feedback. This agile approach will help to ensure that the solution is actually what the business wants – unless they are able to see the solution in action it’s impossible for them to really know what they want! An added bonus is that they will feel that they have shaped the project and will have a vested interest in the outcome. 

Recoup your investment 

From the C-Suite down to the office floor, what everybody wants is something that makes their work easier and the business more successful. If you can demonstrate that what you are doing is going to achieve tangible results they will sit up and listen. For example, a report which currently takes 5 hours will take 1 hour as a result of this project. And if that report is created by 100 people at a cost of £50 per hour, the bill drops from £25,000 to £5,000. 

To hammer that point home, McKinsey carried out analysis over a five year period which showed that companies who put data at the heart of their operation enjoyed marked improvements across all departments, with sales and marketing ROI increasing by 15%-20%. Investing in creating a data driven culture is vital for the growth of any business determined to stay ahead of the pack. 

Modern Data Warehouse – Snowflake & Fivetran

Why AI is a lot like teenage sex – and how you can get better at it

In 2013 Dan Ariely, professor of behavioural economics at Duke University, got the analytics world all aquiver when he stated: “Big data is like teenage sex: everyone talks about it, nobody really knows how to do it, everyone thinks everyone else is doing it, so everyone claims they are doing it.” His risqué comments appeared in the midst of a data revolution, the experience of which was less than satisfactory for many companies.

Six years on and we’re doing it better. However, AI is in danger of being in the same position as Big Data was six years ago. Some data strategists would have you believe that AI and its bedfellow machine learning are better than sex. But as with anything in the data revolution, don’t expect to achieve total fulfilment overnight.

Cost vs kudos

A report from the McKinsey Global Institute predicts that early adopters of this technology could grow the value of their business by 120%. It adds that those who fail to jump on board the AI gravy train could lose a fifth of their cash flow. No surprise that companies are throwing money at the problem – but not necessarily always for the right reasons.

Kudos is not enough of a reward for a business to pump considerable sums of money into an AI project. A survey by the US analytics company Figure Eight showed that the majority of companies were spending at the very least $50,000, rising to over $5 million for those serious about making it a central part of their business.

AI has now been around for two decades. Advances in tech plus heavy investment from the likes of Google mean that the cost of leveraging AI tools will continue to fall, levelling the playing field and allowing smaller businesses to utilise AI. If you’ve made it your business to amass a wealth of clean, properly managed data you are already well positioned to launch an effective AI project.

Let’s not get ahead of ourselves

With any data project, there are things you need to think about before you get started. If you’re looking to get started with AI specifically, first consider whether the problem you would like to address is best served by the technology. Don’t expect AI to act as a sort of panacea; you need to be deploying it in the right way and for the right reasons. If you’re unsure, talk to an expert first (yes, we can help with that) to assess what sort of data analysis would be best for your particular problem.

Maybe there is an area of your business you are certain would benefit from an AI solution – if only you could convince the CFO to invest. If they’re keeping a tight hold of the purse strings ask yourself: does this align with the greater corporate strategy? If no, you’d probably be better focusing your efforts somewhere that does.

Finally, when you have identified the right AI project and hired yourself a crack analytics specialist (hello), don’t assume that the thing will just run itself. AI is smart but it still needs help. That means putting together the right team – and not just a couple of people borrowed from the IT department. Successful AI needs buy-in from people who understand the business need and who are working with the numbers on a daily basis.

Get it right and you’ll transform your AI experience from a meaningless one night stand to a satisfying relationship that grows into something really special.