More and more analysts and companies are looking for analytics solutions that can provide real-time data, the famous real time analytics. They condition the hiring and use of tools only if they offer real-time reporting.
In this article I will show why real-time analysis is overrated. I will also give you tips on when your company will be ready to use them and get results from them.
- Why real-time analysis is not relevant
- When real time analytics is recommended
- In conclusion
Why real-time analysis is not relevant
Those who know me know that I am not a big fan of real time analytics. Obviously this is not the reason why real-time analysis is not important. At least not for most companies.
Reasons why I don’t consider real-time analysis as relevant:
More reports and less analysis
In the web analytics universe we already have so much data to analyze that real-time data only makes things worse. Not that I’m being lazy, but analyzing a large volume of data and having insights that bring value to the business in real time is not as simple as it sounds.
First, any analyst who brings more than one insight per week to your business is cheating you. No one can bring relevant and valuable information to your business every week. It takes analysis and good judgment.
Second, reporting is not analysis and insights. Real-time analysis tools offer several types of reports, but few (or almost none) offer any insight that can be turned into results.
Too much investment in tools and not enough in people
The search for real time analytics causes companies to hire tools and professionals in the wrong way. A good analyst can bring valuable information to your decision making just by using free tools such as Google Analytics, for example.
Most of the time companies take the opposite route, hiring robust tools like Chartbeat and Piano and delegating the analysis of the incoming data to an intern. The chances of getting good reviews are very low.
On the other hand, a good professional is able to get the most out of free tools at a much lower cost.
Read also: Analytics tools: 5 reasons not to use them
Increased complexity in systems and processes
Here is a point that most companies do not pay attention to. Capturing and processing all the data in real time demands much more complex and robust structures and processes.
The collection, processing and storage of this data requires a much larger and better structured infrastructure.
Wrong decision making, but safe
The terrible false notion of trust.
Many think that having the data in real time will make them more prepared for the competition. This is not always true as it can be quite the opposite.
Companies don’t realize the real cost involved to be able to work with real-time data and the return on investment.
So far we have seen reasons why I don’t consider real-time analytics to be that important. However, this is not to say that you can never benefit from real-time data.
If you think your company is ready to analyze real time data, see if it meets the following requirements.
When real time analytics is recommended
Do you get enough visitors that a real-time analysis can represent, with a good level of accuracy, the behavior of a segment? Does your site have enough conversions in real time for any action to be taken about it?
If you have 100 visitors per day on your site and 2 conversions probably real time is not a good fit for your analytics strategy.
This is not to say that you need thousands or millions of visitors to your site, but you do need at least a significant number of visitors with an expected behavior that can serve for analysis.
To analyze data in real time you need to have at least one full-time analyst on staff who has good analytical skills.
A professional with good analytical skills will know how to distinguish relevant information from noise. I work with publishers who delegate real time to copywriters and editors who are unaware of business metrics and KPIs and therefore end up making poor decisions.
Decentralized decision making
If a company wants to adopt real time analytics it needs to empower its employees and decentralize decision making. Real time is now. Not tomorrow or a week from now using real time data from days ago.
Suppose your hierarchical structure requires you to receive the data in real time, then report to your superior, then the superior reports to the supervisor, who, being in agreement, sends it to the CxO to make a decision. What good is real-time data if your suggestion to pause a PPC campaign will take 3 days to happen?
Operational execution capability
You have identified a point of improvement, a banner or a block on the cover of your site, and you want to change it immediately in view of the fact that real time has indicated that it can be improved. Do these campaign and website changes take two or three days to occur? Most likely real time is not necessary in your company.
If you can’t change an ad in a campaign or a banner on your website in a matter of minutes, real time is not for you.
“Data on time, not real time”
As I mentioned earlier, if your decision making takes days, there is no point in investing in real time.
Here is a very common flow: real-time data is analyzed (about 2-3 hours) and presented to the superior (1 hour) who in turn sends it to a decision-making committee (1 day). The committee queues and analyzes priorities (2 weeks), questions the numbers (1 week) and only then sends them to the decision maker. This in turn identifies that the decision should have been made 3 weeks ago and makes a decision based on its instinct within 5 seconds.
Do the math. Did the real-time data have any value to the organization?
Almost all large companies have a complex and often plastered structure and are unable to react in real time. If you can’t react in real-time, why do you need real-time data?
A good analyst will know the right time to present data and insighs into the organization.
Concentrate your efforts on important and strategic things. Real-time analytics costs too much!
Avinash Kaushik proposes a way to check whether you really need real-time data:
A human has to be involved from receiving the data to making the decision? If the answer is yes, then you don’t need real-time data, you need data at the right time. If the answer is no (say you have automated intelligence/rules systems), then you need real-time data.
If your company can fulfill all the above requirements, it is possible that real-time data can make a difference. Otherwise, invest in people and improve the analysis and decision making process.
Real-time analyses are not relevant for the following reasons:
- generate many reports and little analysis;
- too much investment in tools and not enough in people;
- increase complexity in systems and processes;
- lead to a false sense of trust.
Real time analytics is recommended if:
- there is statistical significance;
- counts on professionals with good analytical skills;
- the team is empowered (decentralized);
- there is operational execution capability;
- decisions are made quickly.
And in your company, is everyone hungry for real time analytics?
Continue reading: Why Digital Strategies Fail
Continue reading: 10 SEO Tools Every Website Should Use