How many SKU’s do you forecast? At what granularity do you run your statistical forecast? Do you track forecast accuracy? Do you set inventory targets? Do you allocate most of your time to the big ticket items? Do you calculate safety stock? Do you do any of this in Excel?
When I first started in demand planning, the sales forecast, and all the analytics, ran off excel, the same was true for my Sales and Operations Planning (S&OP) counterpart who was the master planner. As I noted in one of my prior blog posts, In Excel We Trust! Depends…, the company I worked for was a $50M business, which grew to a $130M business. The company needed a quick response to a highly volatile, growing market, and, as you can imagine, Excel was not cutting it. The time required to prepare a reliable Excel plan usually led to missed opportunities, and more critically, lost sales due to the inability to have the demand plan updated in time to drive supply. Because our lead times were 90 to 120 days for raw materials, each day the company postponed driving the demand plan, it negatively affected the supply plan. In addition, since we were planning demand in Excel, the supply side was also being managed in Excel. Inventory sizing, safety stocks, and production plans were all being cobbled into one monster of a spreadsheet.
Once we all agreed on a plan, the precious time that we lost managing the monster started eating into our supply lead times. These delays in turn manifested themselves in late shipments due to components not being available. Customers started to penalize us for our delays and, thereafter, started to cancel orders. Since our lead times were so long, these cancellations were occurring while the raw materials were already in transit or in production, creating huge levels of slow moving and obsolete inventory (SLOB). After the company lost millions on SLOB inventory, the executive team finally agreed to invest in software to help reduce the planning time.
You’re probably thinking at this point, perfect, happy ending, you're getting software! Sadly, the pain had just begun. It was 100% the correct choice to adopt new software, however, selecting the correct one for your business is not a trivial task. Before I begin, I learned quickly it’s not really about the formulas used, most software is very similar in its calculation capabilities—what is important is the software's functionality and flexibility. Our first implementation was extremely costly. While I won’t name the company, it is one of the most popular planning softwares on the market, it required hardware, lengthy implementation time, and above all a price tag close to $500K. So what went wrong?
My number one complaint was its user-friendly functionality. Many promises were made during the sales pitch as far as the capabilities of the software.
While I would not say we were lied to about the software’s features, I will say if you took a rubber band and called it “The Truth,” then stretched the rubber band until it no longer had elasticity, well you get my point….
Many things that were our priority and came across as standard features turned out to be possible only by taking tedious steps and rigging other functions together that were not designed for what it was doing. It was like turning a clothing iron upside down and dropping a NY Strip on it to grill—that is, it will work, but it is clearly not the intended purpose. One year after the software was adopted, and 3 months behind schedule, frustrations peaked, as the lack of functionality led to simply running a forecast, then exporting to Excel to validate and adjust the forecast as needed. The final decision to pull the plug came after the most needed function and most timely (when done manually) could not be delivered as promised.
Supersession (linking history of old product to a new replacement for forecasting) was extremely important to our business, as most of the products had an 8-month lifecycle. It turned out the supersession functionality in the system was to be launched at some point but what we quickly found out was the function had been riddled with bugs and no definitive timeline of the launch could be given. Without the systematic function for supersession inside the software, hundreds of SKUs had to be manually manipulated.. Can you imagine being an “Excel Junkie” and wasting two days on such a tedious manual task, especially after being promised better days ahead with software?
Can you guess what happened during the next forecast cycle? When the software reps felt the standard was met with my tedious labor? Let’s just say my sobriety from using Excel hit a brick wall and I did what any savvy Excel user would do, I exported the data into Excel and rigged up my own template to do the work for me.
After spending $500k and a year of time, the company was still stuck using Excel for complex tasks. In the end, we pulled the plug on Software number 1.
There was a lot to learn from going with the “Big Box Solution”:
- Because the software lacked flexibility, the software consultants in charge of implementation pushed to have the system drive process, as opposed to the process drive the system. In other words, every business and industry is different and has different needs. Critical advice: Be cautious of software consultants who argue a food company and a technology company have the same needs—it’s simply an excuse for lack of versatility.
- When the executive team gets the sales pitch from the software company, it is very high level. This means, if you’re the user, you better ask the consultants to actually show you the software’s features in action.
- Just because it has a hefty price tag does not mean it is the best. These “Big Box Solutions” have a rude awakening for the future if they don’t adapt to the latest technology. These companies may have a great brand and a good book of business, but, unfortunately for them, prospective customers that research software will find out that they can not only get quicker implementation time, ease of use, and more versatility from some of the newer alternatives, but they also come at a fraction of the cost.
- This should be a given, but have a well researched cost-benefit analysis on more than one software option.
- More points to come in the Part 2, but my last point is more opinionated, if the graphics still look like the first ever release of windows… I mean Come On its 2014!!!!
There is much more to this, which I will address in upcoming blog posts. In Part 2, I will tell you a true David and Goliath story and how investor politics can lead to some incredibly insane business decisions.
