Analytics-blog

Recently Boris Evelson with Forrester Research released a report titled “BI In The Cloud? Yes, And On The Ground, Too”.

As part of his research, Boris conducted a survey as to why firms did not engage the services of a SaaS business intelligence provider.  Interestingly enough, most of the concerns were ones that AOD and some of our competitors can easily dispose of.

There were about 10 concerns, but most replies were included in the top six concerns.  Below I will outline the concern and give a sensible reply to those concerns.

Security Concerns

Almost 50% of the respondents cited this concern in their reply.  Speaking for AOD, we are SAS70 II compliant to the point that HIPPA compliant hospitals are among our customers.  There is nothing more private than hospital data.  In addition, our entire business hinges on making sure our one system works well. Your internal IT team likely has multiple systems to manage.  So there is a chance we are more focused on application concerns; including security.

Integration Challenges

Respondents were concerned about the ability to integrate.  Actually, this is an area where AOD excels.  We have many pre-built connectors across all types of systems.  In addition, our team has over a decade of experience working with blue chip companies on extremely complex data integration projects across all types of “traditional” BI platforms and software.

Total Cost Concerns

Based on our aforementioned experience implementing traditional BI solutions, this is not a legitimate concern.  Traditional BI solutions are about 10-20x the cost of our SaaS BI solution and take a long time to implement.  You can read more in our white paper.  Also, if you are questioning the value of Business Intelligence, then please click here and we’ll send you a copy of Aberdeen’s recent report on the subject.

Lack of Customization

Most serious SaaS BI players have addressed this concern with a strong  “Self-service” model.  In the case of AOD, the release of our V3.0 self-service application with drag and drop customization, this concern has been met directly, allowing users to customize on  the fly and save to build internal report libraries.  In addition, our experience has shown us that the biggest challenge with BI initiatives is not delivering functions, it’s adoption by end users.  The beauty of SaaS BI is that you can almost entirely focus your staff on adoption instead of project management and requirements definition.

Application Performance

The great thing about a business intelligence system is that it does not require mission-critical performance.  It just needs to work when you access it and be consistent.  Our system does run quickly and is designed to be available at all times.  In addition, should you decide that time is critical we can certainly set-up the system to be real-time.  More importantly, you want to make sure that your data is accurate and can be depended upon.  Performance issues are rare in the age of high processing power.  Data issues are commonplace.

Difficulty of Installation or Risk of Migration

There’s nothing to install.  Sound good?  In terms of migration, all your data will be standardized or normalized inside of our system.  We can always give you back your data.  In fact, we’ll provide you an active backup just as a failsafe.

As you can see, the scenarios depicted above are not a legitimate reason to abstain from SaaS BI – certainly none justify spending millions on customer built BI when you can have SaaS BI for thousands.  Of course, there ARE some situations where you would be better off with an in-house, custom system and we’ll be the first to inform you of that situation when it occurs.  We have seen from experience other SaaS vendors that drove themselves into the ground trying to be “all things to everyone”.  We believe that we are “most things to most companies” but there are highly specialized scenarios where you might want to consider an in-house solution.

Wow.  That’s quite a story/stat.

Of course web analytics are not perfect as most of them depend on cookie technology that can be easily defeated.

However, it’s good enough for most folks given where they are in their knowledge.  If it’s not good enough for you, I suggest you check out Convertro.

I think we’ve all heard about how statistics can be made to support “any argument you want”.  I think maybe that phrase is wrong…it should say that statistics can be made to support “any argument you want if the recipient isn’t paying attention”.

That’s a fact.  It’s really all about the recipient or the audience.  If they don’t thirst for the truth, then they will find themselves fooled time and time again.

For analytics it’s exactly the same.  You can have the best system in the world, but if you do not have the time to seriously contemplate the numbers, then you probably are better off not spending any time with analytics at all.   Just run with your gut.

After all, drawing the wrong conclusion is sometimes worse than no conclusions….and ignorance is bliss.  In addition, for a lot of people, companies etc…inertia is the ONLY force.  So you can join that crowd.  Enjoy.  Don’t call me.

But if you are the type that challenges numbers and understands that the real work begins AFTER reading a graph, chart or table, then by all means you should invest in some sort of analytics or business intelligence system.

Of course, it is equally important that you approach your analysis with discipline and consistency.  Speaking personally, it’s the consistency that is the struggle for me.  It’s very tempting to go down one path, learn a little bit, get distracted….and suddenly end up somewhere else.  Obviously, you need to clean you plate before asking for seconds.

So that’s the trick with analytics.  You must understand the challenge at hand and have the discipline and consistency to utilize your tools effectively.  (or hire someone to help you)

There is no analytics system in the world that can “tell” you what to do.  (If you have such a system in your control, it’s actually not under your control and you are probably headed for bankruptcy or imminent disaster at some point)

We are in the process of turning a pumpkin into a…uh…better website.  So keep your eyes on analytixondemand.com.  It’s about time that our website matches the quality of our analytics offering!

I’m going to speak out today against the incessant march toward simplicity.

YES – it’s good to start simple…especially when it comes to analytics.

NO – that’s not where you should stop.  ESPECIALLY if you are in management.

I read a post today about various “brilliant” people talking about human nature and applying those learnings toward how to manage a business.

The end goal was “let’s dumb it down”.

Really?

I seem to recall thinking that my schooling was “sufficient” sometime around the first grade.  Thankfully, some adults around me thought differently.  As it turns out, it was a good thing they kept me going.  ”Graduated with honors from 1st grade” just doesn’t pay well.  (unless you work in Hollywood)

And that’s what I’m saying right now.  If you don’t like thinking and if you don’t want to challenge yourself…then I would suggest you remove yourself from the business world before it is done for you.

Enough of this catering to the undisciplined.  Most business intelligence software and analytics software providers offer dashboards and KPI’s.  But they are just a starting point for analysis.

We also provide tools that allow for more detailed analysis.

Use them.  Please.

The folks at Gartner see a bright future for business intelligence.  It’s going to be embedded in everything- it appears- including social networking.

Do they see it making you a burrito?  Unfortunately no.  And that’s a shame because I really like a good burrito.  (especially for breakfast)  I also think the term “BI Burrito” has an nice ring to it.

Now back to the topic at hand.  Is Gartner alone in their burrito-less vision?

Nope.  The CTO at LogiXML also seems to buy into this vision.

And so do I.

But frankly, I still don’t see many companies using business intelligence in an efficient manner.  It’s difficult to me to contemplate all the other applications for business intelligence when the corporate market is STILL learning how to utilize BI across the enterprise.

In addition, I’m not sure what the revenue model looks like for a social media business intelligence application provider.  But, I do know that there are serious costs involved in providing any sort of analytics or business intelligence.

So yes, I’m optimistic about the future of analytics and business intelligence…but I’m not sold that the future has arrived in any meaningful way that can impact the bottom line.  And, I’m sure not interested in the “build an audience and somehow you will monetize it” approach.

Those days are at an end with all the financial issues out there.

So I guess we’ll see how long it takes before the future arrives.

I want to continue to address the concept of making sure you mix qualitative in with the quantitative when making decisions.

Suppose you worked for a large company that rolled out a product in their home country that was a massive flop.   Suppose you were not the only failure in that industry…instead everyone else who tried to bring that particular business model to life also flailed.

Would you stick with it or give up?

If you just looked at the sales numbers, reviewed all the various doomsday reports about the lack of growth projected for the industry and sifted through the negative news that resulted from the previous failure….then you would likely want to give up.

In my direct experience, that would be a many-billion dollar mistake.

In spite of all the negative press and sales numbers, your problem was not a product/service that no one wanted.  Your problem was a marketing problem.  Wrong price.  Wrong message.  Wrong distribution…but the right product.

This is a situation where your business intelligence system is perfectly useless.  It’s rare, but it happens.  As I stated in the previous post, this is where the quantitative can let you down.

Empirical data can be useless when you are trying to start a revolution.  Something to keep in mind.

Can business intelligence software and analytics replace human beings?  In other words, can we automate decision making using algorithms.

Well…let’s talk about the financial industry and their “success” with automating their decisions.

Long Term Capital Management.  Oops.

The entire mortgage securitization business.  Massive Oops.

What went wrong?  In all cases they mistook trends for truth.  Say it out loud with me….”Past performance is no guarantee of future results”.

And that’s it.  They forgot the human element.  They forgot that humans say one thing and mean another.  They are emotional.

Those algorithms just don’t have a lick of common sense.

Now I’m a big fan of Ray Kurzweil and his various predictions regarding the merger of man and machine.  His track record is solid.  He thinks that over time mankind and machines will merge.

Now maybe at some point humans/machines/borgs will someday think in a more logical fashion.  Then we can eliminate all the manias, the panics…all the emotion that accompanies human nature.

Sounds boring as hell.

However, for now we need to understand that while business intelligence tools and complex statistical formulas can help inform our decisions, they cannot make decisions for us.

It’s still critical to get out there and understand socioeconomic trends that lie outside of your business intelligence system.

For example:

-Talk to your customers.

-Talk to your non-customers.

-Study macroeconomic indicators.

-Read a book or two that have different points of view on the same topic.

-Spend a little less time listening to the hyperbole of the major news networks.

Then you can see all sides of the story and have a better chance of making the optimal decision.

Better chance than any robot at least.

IBM recently interviewed over 2,500 CIO’s to learn how their roles are evolving, and what they view as important. It’s an interesting report and an easy read.

So here’s one of the stunning results: 83% of those surveyed stated that business intelligence was their top priority.

83%! Seriously…have 83% of people ever agreed on anything? This is either a great sign or a sign of Groupthink Gone Wild ™.

I’ll side with the former and say that these C-level folks do have a clue about what is important.

Second on their list was virtualization at 76%. Virtualization came in second? Is anyone else surprised by this? Virtualization is primairily about cutting costs. Business Intelligence is about decision making. This is a big deal for CIO’s. Now, CIO’s are thinking strategically instead of thinking just in very simple terms of how to cut costs. (which was one of the major points made by IBM in the report)

Now the reality is that business intelligence, when properly utilized, will lead the CIO’s, and their minions, to make better decisions. And that will likely save them more $$$ than any virtualization exercise.

Nonetheless, it’s a little stunning to see the C-level thinking a little more on the abstract level…especially given the economic circumstances.

My day has been made.

Because I said so.

Oh that doesn’t work for you? Well that’s a shame. I just tried to spare you a page of reading.

But…since you are a hardheaded “logical” type person read below and knock yourself out.

Almost every day I see a discussion on the social networks where someone posts “what is the difference between SaaS Business Intelligence and regular ol’ Business Intelligence?”

After I got past the initial shock that the entire world “doesn’t just get it,” I realized it’s a pretty legitimate question.

So I’ve compiled a short list of the “Big Six” differences between SaaS business intelligence and traditional business intelligence.

1.) No hardware required: SaaS means you don’t host the software within your company. That means you don’t need to buy any servers, procure any cold rooms, or talk to anyone in IT about anything hardware related.

2.) No software integration expertise required: You don’t need to hire a consultant or staff up with “expert” business intelligence employees. Instead, you just need a web browser with a web connection.

3.) Requirements Document: You can throw it away. You can cancel your committee meeting to help write that document. 90% of what you need is likely in the SaaS system already. The other 10% will likely be in the system by the time you’ve mastered the first 90%.

4.) IT: You may not need them at all. It just depends where your data resides. But I can tell you that you probably don’t need very much of their time.

5.) Cost: SaaS vendors charge thousands per month with very few upfront costs. Traditional business intelligence requires a substantial capital investment up front, and sometimes their solutions end up costing millions.

6.) And finally, the most important thing here is SPEED: So the SaaS solution just eliminated most of your discussions with finance, eliminated any hiring decisions, reduced or eliminated time spent with IT…. and will actually deliver reports and analytics you can use within weeks.

It’s awesome. Just like I said…one page earlier….