the attractive but un-sexy side of BI

I had a wonderful opportunity recently to join some very smart colleagues on a BI (Business Intelligence) panel presentation. I first need to thank Jon Inge for inviting me to be on the panel and for being an advocate for the un-sexy side of BI. Second I need to thank Rick Heuer, Geert Mol, and Clive Perry-Core who were on the panel with me. I learned a ton about CRM BI from Rick and Geert, and Clive has more hands-on experience to the nuts and bolts of BI than anyone I've ever met. It ended up as more of a round-table discussion than a panel really, but it was great for my own learning opportunities that way.

In his opening Jon referenced a great article from the Wall Street Journal on "Knowledge Management" which is largely what BI is all about. (Which I was bummed to not find on the web and provide as a link to all of you. If you find it give me the link in a comment.) The opportunity benefit of BI is to effectively act on meaningful information about your business...or more briefly put knowledge management.

During the introduction of the panel each panelists described how they worked in BI. During this brief introduction I came to the sad realization that I specialize in "un-sexy" BI. The "sexy" side of BI is in CRM and marketing. Their personal introductions were exciting and I'm sure led to everyone dreaming about a hotel knowing you when you walk in the door, welcome you by name, hand you your favorite margarita, quickly tele-port you to your room with pre-set temperature and lighting preferences, and have music playing based on your perceived mood and core body temperature. It is what you read about in press releases from the big brands in the industry where they know everything about what the customer prefers. The summation is Know your guest and achieve total patron value! Increase revenue through understanding your guest's preferences and market directly to them! Sexy No? (to be said in a French accent)

For the record I totally agree it is way more fun to talk and dream about the benefits of BI with CRM. But the reality is the ROI and TCO for BI with CRM is harder to quantify (I tried hard to get a fifth acronym in that sentence but I couldn't find one that made sense. Though I digress). It is not impossible to quantify, just harder. Before I jump into the "attractive but un-sexy" side of BI I think it is important to lay some foundation. First let's take a shot at defining TCO. Total Cost of Ownership (TCO) is the total investment capitol put into the technology solution over a specified period of time. Return on Investment (ROI) is the duration it takes for decrease in costs and/or increase in revenues to surpass the overall investment (or TCO). To put simply, save or make more money than what you spent within a specific time period. The measure of any good technology solution should be create cheaper, faster, and better operations than what you do now. The general problem with proving the ROI of BI with CRM is the generalizations that take place into why certain customer preferences exist and under what specific set of circumstances. It is the accurate capture of circumstances that really creates the challenge. It is difficult (not impossible) to tie a decrease in cost or more likely and increase in revenue to a specific preference or area of guest behavior being influenced by an almost infinite number of inputs (aka circumstances) that lead to a specific action (for more thoughts on this see when CRM will achieve its potential).

So if BI with CRM and market data is "sexy" then the operational analysis side is really the "un-sexy". Note that both are attractive, but the former is far more attractive than the latter.

The un-sexy side is the analysis of your operational systems data in order to decrease costs and increase revenues (yawn). I'll define operational systems as Property Management System (PMS), Point of Sale (POS), inventory management, time and attendance, service optimization, and other solutions implemented to manage the operations of a venue. See, I'm only two sentences into the "un-sexy" and you're mind is already wandering...hmmm..what should I have for lunch today...

But, when this data is captured in a meaningful way and linked to other external data elements (i.e. weather, currency exchange rate, website click through, etc.) the value proposition of BI becomes very attractive. I'm going to say "becomes" attractive because generally the first thing that a BI system shows you is how poorly your systems are currently setup to capture meaningful data. You will need to spend at least three to six months cleaning up configuration elements of your existing system(s) before BI will begin to have analytical payoff. The correlation can be made that it's like exercise for your technology solutions. You may hate the process but you'll like the results. I'll cover some thoughts on the steps to a successful BI installation in a future post. But for those in need of immediate gratification Jon Inge wrote a great article in this months Hospitality Upgrade (it's free by the way...so subscribe) entitled Demystifying Business Intelligence. You can download the article here from Jon's website.

As an exercise in brain storming here are a few examples of how to both define and achieve ROI for your BI solution. I am making a HUGE assumption here that you already have some meaningful KPI's identified for your business and that you can apply them accordingly. If you have not, first go read Jon's article.

the pms story
A key metric for any hotel/resort is RevPAR (revenue per available room). The value of this KPI becomes more valuable over time as you compare it against historical data. It is one of the measures used for how to price a room and it's amenities. Some PMS systems will give you this number over a specified date range and may let you compare it to one other range. Or you can run multiple ranges and do the comparison yourself by entering the information into a spreadsheet. But, what if rather than just looking at previous RevPAR numbers you were able to review historical RevPAR numbers with the corresponding temperature/weather and current exchange rates with selected currencies over multiple selected date ranges. Then what if that information could be plotted graphically for you giving you some solid predictive analytics on what room rate would give you the best RevPAR based on matching factors from a historical period. The ROI benefit in this example is actually achieved in more than just the revenue opportunities. It can also be achieved in the "improved costs" column by spending fewer hours entering information into spreadsheets.

the pos story
The POS story I think is one often left behind, but it is one near-and-dear to my heart. Your POS (and PMS) system holds real data about customers that they pay you to keep so you can know more about them. How wonderful is that?! Grocery stores have the POS BI story down...it's a science of revenue for them. Hotels, resorts, and other venues should be watching and learning from their successes and their mistakes. As a quick story lets meander into a world where all your POS data means something (meaning versus what you have now). It means what do your customers like, what do they not like, and what should they like but just are not finding and buying. By "should they like" I'm referring to placement, description of the item, presentation on the plate, verbal marketing, and price. BI with POS allows a manager or chef to aggregate and compare data for items that are selling well to items that are not selling well, and possibly against the day of the week, time of the day, and temperature outside.

the inventory management story
You get an even better story if you tie POS to Inventory Management and look at actual cost versus selling price data. A manager can run pricing strategies and get concrete data back on how well or poorly it worked. For example, take your top 10 selling entree items and increase the price of each of them by 5%. If you have historical price and cost for those items for the previous three months you should have food costs, theoretical and actual yields, margin, and profit for those items. Track on a weekly basis for three months how the 5% price increase affects each of those performance monitors and you begin to understand exactly how much items should be priced for in order to achieve minimum food costs, maximum margins, and maximum profits. You may tell me you can do this now, but I'm going to tell you only with a small troop of analysts and well massaged data.

the employee performance story
Wouldn't it be great to have a better understanding of what made great employees great, good employees good, and poor employees...well...suck? I'm sure you can tell me about personal drive and character and customer relation skills and I'll agree with you. But what if you could see metrics about employees like how many more transactions they perform, or how efficient they are in certain work areas, or why they are always able to predict what needs to be done before it needs to be done (okay...that one may be harder to figure out). Again, manufacturing plants have been dong this for years...we just have to catch up and figure out how to make it work for hospitality A BI solution should have all the data points where an employee enters information into your technology solutions, and that transaction is tied to an employee ID. Employee ID is often the ONLY common thread among disparate technology solutions so this specific area can be one of the easiest to achieve. The benefit of seeing performance information on your performance range of employees is to spot operational efficiencies that great employees implement and then work to train those into the rest of the staff.

If you are shaking your head at me saying, "Employees won't do it even if I show them how to be better", then I'm going to tell you that you have bigger problems. Employees suck either because of lack of education or apathy. BI can help you fix the education part and then you can help yourself and fire the apathy part.

the staff management story
The last story we'll weave is the magic tricks around staff management. I never ceased to be amazed at the creative processes that I find managers and client venues using to estimate staffing requirements and build employee schedules. It is an artful guessing game made more accurate after years of experience and some trial and error. But if you have employees entering information for time and attendance (if you pay people you have this I think), and you have historical revenue information (for a specific day of the year possibly for multiple years), and you have employee performance detail (for example revenue per hour) for a "good" employee then you could have the best staff management procedures on the planet based on BI data and predictive modeling. Bringing this information together into a BI solution gives you forecasted revenue expectations with forecasted costs for which you can then effectively manage your schedule and get maximum revenue potential at minimum cost from your staff. I have a dream....

See, you find it attractive...admit it. But (yawn) sooo un-sexy. Who wants to talk about boring operational efficiency metrics to executive management so you can purchase a BI solution? I hope you.

I would love to hear from you if you have other great examples of ways you achieved the ROI of your BI solution. I'd also like to hear from you if you think mine are ridiculous. And as always, if you'd like to find out more about kevin sturm Consulting please visit my website or email me.