Looking forward at the Analytics Space in 2017
We’re 1.5 months into 2017 and I think I’m ready to predict what’s to come in the analytics space. I did a similar (albeit earlier in the year) exercise for 2016, which can be found here. I think many of these themes are still true, but they perhaps have a longer trajectory than a statement of change in 2016.
Now, what is going to take place this year in Analytics? I think there are 5 continuing themes for the year and I’ll do my best to steer clear of the over played themes like AI, MACHINE LEARNING, and IOT. (Disclaimer: I think these are all interesting, disruptive and happening in all sorts of places. However, the majority of those who will read this will not be enlightened by me proclaiming “AI is here to stay” or “Machine learning is taking over your coffee maker” or wait…)
Anyways, onwards to those 5 things in analytics that just might happen this year.
Someone, somewhere will make it easier to combine clickstream data and revenue / sales related data.
I’ve spent time with digital marketing groups at some large companies and I continue to see the challenge of combining clickstream and sales data. By clickstream data I mean web activity data (often provided by Google Analytics, Adobe or a myriad of other small tech companies). Combining this with customer sales related data is the challenge. [Caveat: most born of the web retailers are doing this well, because they invested in it. However, I haven’t seen many enterprise software companies do this well, yet.]
Some of the above web usage / analytics companies will tell you that they provide an easy to use, integrated view of your customer journey. FALSE. Those products likely provide detailed usage / activity on your various web properties, and perhaps even allow you to upload some sales related data. However, they are not providing a solution for navigating complex relationships between web visitors, paying customers, returning customers, etc. This is especially true when those visitors continue to use non-tracking software or flush the cache.
That’s enough backstory. I think that either one of the big players or an up and comer will make this abundantly easier. Especially for marketing groups that lack the technical abilities to create their own custom solution OR don’t have a fruitful relationship with their IT counterparts.
As many market analysts have predicted, I too think there will be more market consolidation in Data and Analytics.
I thought companies like Tableau would get purchased last year when they were at a nice discount. That may no longer be true, but I still think they are a great target for some of the legacy players or new(er) cloud companies.
I don’t think data prep companies can live a long life on their own, so they are also a prime target this year for either Single Discipline Toolsets (IE Data Viz, Data Governance) or mature ETL companies.
Lastly, I think many of the “easy to connect to your cloud data source” startups and earlier stage companies will falter over the next couple years. I talked about this in last years predictions, but believe it more fervently today. Unless they UP their value proposition, they won’t last.
Google plays catch up to AWS and Microsoft Azure in the enterprise cloud race, but they will catch up soon(ish).
A wise individual recently told me that Google’s foray into the cloud platform space has been akin to Microsoft 90’s strategy re: browsers. I thought this was really interesting and seems to align.
They have been a bit slower on the uptake here, compared to their large Seattle competitors. However, their innovations with Google Cloud Platform (GCP) and the underlying data technologies and libraries is exciting. Now, the Google ecosystem is still maturing. If you’re a company looking to leverage GCP for your app or data platforms, just know that you may be one of the earlier adopters. (IE not a ton of documentation, not a ton of similar companies that have tried with success or failures, and not as much talent in the marketplace.)
Lastly, who doesn’t like a 3 team race? It’s better for everyone really. I’m not rooting here for anyone in particular, but I do like the idea that we’ll see more frequent innovations, lower costs and options to deliver the next [insert buzzword] tech.
The quality of data and how it’s governed [managed, integrated, analyzed and reported on] is still a massive challenge for just about everyone but tools will hopefully make it easier!
This is the un-sexy part of data and analytics and it continues to be a massive challenge for everyone. I could explain why that is, but I think you know why.
So Kyle, how do we solve it? Not so fast. I’m not here to solve your problems (I have a day job for that) but I’m here to predict, perhaps, what may happen this year that could help. I wrote last year about companies like Alation, which are tackling this problem head on with their easy to use tool. I still like what they are doing.
I think a larger theme here is that this will become more integrated into your data tools. The more mature data companies (think Oracle, SAP or IBM) all have created tools that help with this. However, they are mostly separate tools. Who wants that? The visibility and transparency for developers, analysts and end users needs to be integrated into their creation or consumption flow.
You might still be asking, WHY HASN’T ANYONE PRIORITIZED THIS? Here’s a fun activity for you, go ask your new data scientists if they want to document their data preparation process or clean up data from your Salesforce platform. Once you’ve done that, go ask your Director of Marketing and your Director of Sales Ops their definitions for MRR (Monthly Recurring Revenue). Report back in the comments on what you learn.
Experimentation will increasingly become adopted with proliferation of hypothesis driven analytics projects.
Companies like Google have integrated this philosophy into many aspects of their business, however it is not as common at many other organizations. With the influx of data scientists trying to use advanced statistical techniques to glean insights for their businesses, there is a need and a surge of experimentation approaches.
Generally, this will start with a business hypothesis such as “We believe we can increase qualified leads in our healthcare segment by 25% in Q1, compared to the prior 2 years.”
From there a series of activities will take place, but the important component here is that companies are starting with a strategic statement and proving or disproving that statement, with data. Ideally, it will lead to some real-world experiments like “We are going to remove all discounts from store x for 1 month to see if our premise for discounting and sales are accurate”.
This isn’t necessarily new as you’re probably thinking, but it will become more common in the business setting this year.
No prediction on companies to be on the lookout for this year…
Unlike last year, I won’t be adding a section on analytics companies to be paying attention to this year. It’s not for lack of interesting companies in the market, but I’ve yet to be thoroughly intrigued. It’s still early in 2017 and I know there are a number of yet to be revealed startups in the space, so more to come here. I do plan on posting in the next few months on some of these new companies.
That’s it. My predictions for 2017 and I only used AI, Machine Learning and IOT in the intro and conclusion. Enjoy the next 10.5 months.
