Do you really need to BUILD a Business Intelligence system in house?

During my extensive career in Business Intelligence (BI), I saw many companies starting to get serious about being a Data-Driven company when they are about 50 employees strong.

Why around that milestone?

  1. Because the data starts to matter even more

  2. Because they have the resources to process the data

  3. Because they have the resources to action the insights they get from data

But could it be that the company would grow much faster and double or even triple the revenues in half the time if that same company (let’s say a startup) had the access to data and insights sooner?

Many founders understand that Data is important, but they think they must invest heavily to develop Data Analytics in-house.

Data Analytics (and Business Intelligence) are definitely important, but let me remind you about the famous 80/20 rule:

The majority (80%) of the results are almost always caused by only a few (20%) inputs.

In other words, 80% of the value that your company gets from the insights are caused by 20% of the efforts to get those insights.

Please understand me correctly. I am not saying that 80% of the work that is invested in Data Analytics is of little value and can be discarded because it contributes to only 20% of the value. I am saying that your company will waste 80% of the effort to get that additional 20% of value if you undertook the long and painful road of building a Business Intelligence system all by yourself.

I am also saying that unless your company offers Data Analytics as a key competency and unless…

  • combining

  • managing

  • storing

  • securing

  • and manipulating

… data is your core business, investing valuable resources into these activities may not be the best way to use them. Getting a trusted partner whose only focus is on creating seamless Data Analytics would be the smarter way to go. So your resources can focus on the 20% of work that drives 80% of value for your company! Because your trusted partner is focused on one part and one part only, they will likely be a lot better and more efficient than you at running your Business Intelligence and Data Analytics.

Let’s say you are running a platform, a webshop or a service company. In the ideal scenario, you want to invest all your available resources to improve

  1. product or service

  2. customer experience interfacing with that product or service

Is hiring a programmer to write code that brings all of your data together is making your product or service better?

Is it making a customer experience interacting with that product or service better?

Indirectly it does” – you may reply

But the same can be said about an email service to contact your customers, however, you do not hire programmers and developers to create for you email programs and you do not hire administrators to support your own email server. Most companies simply get Google to do it for them by subscribing to Google Apps for Business or by purchasing Microsoft Office with the baked-in Outlook.

Keep in mind that: it is what you choose to do with the capabilities that the solution enables you to do, that matters!

One of my favourite statements is from a Harvard Business School Professor Theodore Levitt states the following:

People don’t want to buy a quarter-inch drill. They want a quarter-inch hole!

So why would you sink over 100K and months of time to build a system when all you need is accurate, timely, consistent insights to help you make smart decisions?

Let’s look at it from another perspective. Running a successful business intelligence (BI) system requires a lot of things to work smoothly together.

A BI System needs:

  1. A scalable Data Warehouse to store all your data

  2. Robust Data modelling with ETL/ELT instance to combine your data

  3. Integrations to bring your data together in one place

  4. Reporting to visualize data in easy to understand reports

  5. Data Validation to ensure accuracy, consistency and timeliness of data

  6. Data Governance to ensure your data is safe, compliant and accessible

  7. Master Data Management to ensure everyone in the company understands what data is available and how to use it

  8. Disaster Recovery to ensure stability and availability of data

  9. Models and Formulas that are unique to each business and which need to be applied to the data

With that in mind, do you really want your team to waste time on:

  • Databases

  • Security

  • Data accuracy

  • Scaling of databases

  • Logs

  • Errors from dozens of 3rd party APIs that you use to generate data

  • Disaster recovery

  • Speed of data analysis

  • Possible issues with data duplication

  • Absence of some data because 3rd party API is unavailable

These are not the activities that add value to your product or service! Data Analysts do. When they analyse your data and give you insights. But anything that happens before that is 80% of work that really gives 20% of the value (at best) if you choose to do it yourself.

Reflecting on what Theodore Levitt had said, what you likely really want is

ALL your data, together in one place

ALL your business logic applied to it

AND timely and accurate REPORTS to help you make the right and smart decisions

Let me give you one more example:

Do you advertise in Google Ads, Taboola, Outbrain, Facebook, Linkedin?

Let’s pretend that you do

What you likely want is to have a good understanding of your Conversion Funnel in each of your advertising channels, plus an Overall Conversion Funnel across to show total conversion across all advertising channels.

Or maybe you want a nice and easy to understand comparison report that tells you which channel is the best for attracting the best customers? It is called Attribution Modeling

I have news for you!

Almost every customer I have worked with, from the Fortune 100 companies to startups with only 2 employees advertise at least on Facebook and often on 1 or 2 other channels. But usually more than 4 channels. The process of getting the data out of Facebook and Google and combining it for a Marketing Efficiency Report has already been perfected by people and teams who did this 100 or even 300 times. (I will probably be on the top 10% tier of this range)

Do you really want to build from scratch what others have already perfected through relentless focus and specialization? And if your answer is YES to the previous question, do you really think you can do it better, faster and cheaper than the people who did it 250 times before you and made many mistakes and learned from them? If your answer is still YES, all business logic would dictate that you should re-purpose your company to be a data integration and analytics enterprise.

My #1 recommendation to founders always is

Always keep detailed data because it will become your strategic advantage as your company grows

But managing that data and everything around it is a huge undertaking if needs to be done right.

It is always better to get an experienced partner who can help you manage your data better, faster and cheaper, so you can focus your precious resources on doing what you do best – making your customers happy!

Reach out to me for more thoughts and ideas on how to build the best Data analytics system.


Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>