Trends in business intelligence software: Augmented analytics and cloud
Computing research: How is BI changing and who are the leaders?
BI was already moving away from the data centre and into the cloud, but the pandemic has put a rocket under that previously leisurely process. After all who needs on-prem when there's no prem to get on at the moment?
In our regular look at BI systems, we asked 200 UK IT leaders and operatives across all sizes and sectors about the solutions they knew best, and compared their answers with those from 2019 to see what might have changed.
In perhaps the most striking statistic, those who said their primary BI solution was maintained on-premises dropped from 47 per cent in August 2019 to 26 per cent in January 2019. Hosted solutions were also down from nine to five per cent.
That's a pretty big change in a small amount of time.
Is your primary BI deployment on-premises, SaaS, hybrid cloud or hosted by a partner?
We also found the number preferring the on-premises option for BI had dropped from 32 per cent in 2019 to 19 per cent in 2021.
Which is your preferred operating environment for BI (2021)?
SaaS and hybrid cloud are stepping in to fill the aspiration gap left by in-house systems, with almost a half (46 per cent, up from 38 per cent in 2019) now preferring pure cloud and 35 per cent (up from 30 per cent) opting for hybrid cloud as their ideal scenario.
Another trend is the arrival of augmented analytics. Pretty much all the big vendors now offer this feature in one form or another.
Augmented analytics uses AI behind the scenes to help the user drill down into data. For example, a single data point such as ‘Sales July' tells us nothing about the variances in sales from different outlets, whether the figures are rising and falling or how they might be correlated with other data.
Instead of having to hire the services of an analyst to grind through all this data, you just type in natural language query or point and click on a chart, and behind the system will select an appropriate machine learning model to deliver meaningful, contextual answers - and nice graphics too - without the user needing to understand either code or statistics.
Augmented analytics represents the latest phase of self-service, which has been the direction of travel for all BI tools for some time.
This democratisation of analytics tools has led to increased emphasis on the front end, with a variety of dashboards and other data visualisation tools now an essential part of every modern BI system. This is where a lot of the development focus has been going in recent years. After all, a pretty face makes it much easier to pull in the punters.
Another knock-on effect of democratisation is that role-based and individual-based security is now a requirement rather than a nice-to-have - it is vital that sensitive personal or financial data is restricted so that not everybody can interrogate or download it.
How important is it that a BI deployment is available as a mobile app? (1 = unimportant; 7 = very important)
And of course BI has gone mobile. Having said that, the importance of mobile apps decreased a little since 2019. This may be because the mobile form factor is limiting when it comes to querying at any great depth; and of course mobile workforces spent vastly less time on the road in 2020/21.
The BI tools market
Other vendors mentioned were: GoodData, Logi Analytics, Microstrategy, ThoughtSpot, Yellowfin, Kognitio, Datameer, Looker, Pentaho, Pyramid, SAS and Targit.
The BI market is dominated by a 600 lb gorilla. Interestingly this gorilla is only 6 years old.
Sixty per cent of our respondents had trialled Microsoft Power BI, and 42 per cent had taken it to production.
Next in line was Tableau, with 15 per cent trialling and 6 per cent in production; and then SAP which has a number of BI solutions on the market including Crystal Reports, Business Objects and SAP Analytics Cloud.
Let's take a look at that top three.
Microsoft Power BI
Thanks to being an option in Office 365, Power BI has quickly become the most widely used general purpose BI suite, not bad for something that's only been around for six years and which started life as little more than a glorified Excel widget.
Although there's an on-premises version, Power BI is heavily pushed as a cloud service.
Power BI pulls in data from in-house and cloud sources. There are built-in connectors to import data from everything Microsoft (and Microsoft recently integrated Power BI with Teams) as well as Marketo, Salesforce and Google Analytics and dozens of other sources. Access control and sharing within Power BI can be managed via Active Directory groups.
There are a few options on the augmented analytics front. Last October Microsoft introduced Smart Narrative, an AI feature to identify key takeaways, illuminate trends and edit the language and format for a specific audience; and Premium users can make use of the AI Insights tools in Azure for options like image tagging and sentiment analysis.
Power BI Smart Narrative. Image source: Microsoft
What did respondents like about Microsoft Power BI? Well first they liked the price - it is free with Enterprise Office 365 and relatively inexpensive if you need to pay for it.
They also liked the range of integrations out of the box, and the familiar Microsoft look and feel meant new users could get going quickly.
Certainly, this behemoth takes some beating, but as a general-purpose BI solution it is lacking in some key areas, respondents said.
The natural language capabilities weren't as good as some rivals', and some respondents objected to Microsoft's insistence that everything should be in the cloud (even though on-prem is available). Then there were the restrictions: for example, every member of staff who wants to view a report stored on SharePoint needs to have a Power BI Pro licence.
Tableau
Second most trialled was Tableau - which is now part of the ever-growing Salesforce empire. Salesforce also acquired Slack last year, so we anticipate some native integrations between the two down the line, a little like Microsoft is doing with Teams.
Best known for its slick UI, Tableau is available in on-premises and cloud versions. In terms of augmented analytics, Tableau offers Explain Data and Ask Data, AI-powered features that allow users to interrogate data points with mouse click or typed query.
Tableau Explain Data. Image source: Tableau
Late last year, Salesforce announced it is merging Tableau with Salesforce Einstein analytics, with the resulting augmented analytics product to be known as Tableau CRM, part of the Salesforce CRM suite.
Tableau users can also deploy the GigaSpaces real-time analytics and machine learning platform.
Tableau's ease-of-use, friendly interface and numerous integrations were frequently mentioned themes.
"It's a complete solution that's very easy to produce beautiful reports", said one chief architect.
On the other hand, some said Tableau is a little lacking on the data preparation side expecting the data to be in a fairly consumable format from the off. Others complained about the scalability and performance problems when on-prem installations grow large, and the role-based authentication could maybe do with more work, one respondent commented.
SAP
SAP has a number of BI and related offerings including Business Objects, Analytics Cloud, Lumira, Crystal Reports and the datawarehouse BW/4HANA.
SAP Analytics Cloud is an end-to-end SaaS analytics platform built on SAP HANA Cloud Platform (HCP), which includes BI along with modules covering risk, compliance, predictive analytics and planning.
SAP Analytics Cloud. Image source: SAP
For years SAP has focused on moving enterprise customers to the cloud, with mixed success, but the latest version of Business Objects offers hybrid cloud connectivity to Analytics Cloud, in acknowledgment, perhaps, that cloud is not for everyone everywhere and that customers need more time to consider their options. Business Objects support is available until 2027.
The company's augmented analytics offering is called Smart Insights, and as of a few months ago the feature can now be used on live HANA data sources.
It's difficult to lay out simple pros and cons for SAP because our respondents were using a range of different products, but to summarise: Business Objects is seen as solid but long-in-the-tooth, whereas Analytics Cloud has been generally well-received and is considered reasonably priced - although with pricing the devil's often in the details.
To conclude
BI is moving to the cloud and augmented analytics is the new hook all the vendors are using to try and pull you in. It's all about the front end, enhanced self-service, and integrating natively to as many different sources as possible. And Microsoft is eating the world again.
Author spotlight
John Leonard
More from John Leonard
Ofcom fines TikTok £1.9m for failure to provide child safety information