Recently I read an article entitled “How to Make Sure Your ERP Vendor Doesn’t Drag Your Good Name Through the Mud” by Adam Bluemner at KPI Linkages.jpg

It’s an interesting article, for sure. And it brings up some valid points.

Nevertheless, I’d go beyond what Bluemner proposes. Here’s my take on a few of the items:

Filter out providers who aren’t focused on your needs

Let’s add some reality to this statement. If the people the vendor sends your way before the sale are unable to make significant and cogent observations about your business, your processes, and your situation in your industry that go well beyond how the technology their peddling might improve things, then chances are the vendor is a “technology installer” and not a real “solution provider.”

My recommendation would be to look for some real love coming from the vendor, in terms of valuable, free advice about how your company can start making more money tomorrow, regardless of what technology or vendor might be selected. Tim Sanders, in his great book Love Is the Killer App, talks about how it pays to give people free, valuable advice—advice or recommendations that will really help them improve and grow—even if you don’t have a business relationship with them. If the vendor isn’t doing this from day-one, what makes you think they are really, sincerely focused on your company’s needs?

Insist on a detailed ROI justification

Here I would add this: do not settle for rule-of-thumb calculations. Get the vendor’s knowledgeable people to sit down together with a cross-functional team from your organization. Have them create estimates on three fronts based on the proposed new technologies and implementations:

  1. INCREASES in Throughput (T)—That is, revenues less only truly variable costs: This eliminates questionable allocations of expenses and the numbers will more closely reflect actual changes in cash-flow
  2. CHANGES in Investment (I)—How many dollars will the enterprise tie up in software, hardware and services to get the new technology delivering the planned increases in Throughput? Will inventories go up or down? What other investments might be required or, better yet, delayed or eliminated in the future?
  3. CHANGES in Operating Expenses (OE)—Including support and maintenance expenses on the new technology. However, do not include any pro-rated or rule-of-thumb “savings” on operating expense that are not backed by determinations to slash real payrolls or remove other hard expenses from the monthly financials.

Each of the proposed changes to T, I or OE should be backed up (on paper) with statements along these lines:

  • Because the new software will provide X, sales and marketing expect an increase of 15 percent in T from market segment A over the three years beginning MM/DD/YY. This will add $N in T in year 1, $P in year 2, and $Q in year three.
  • Because the new software will allow Y, we do not expect to have to open a new branch location in YYYY in <City>. This will reduce forecast Investment by $Z, and reduce forecast OE by $D over three years beginning MM/DD/YY.


All changes to T, I and OE should be directly linked to a benefit derived from the proposed implementation. The numbers need not be accurate to the Nth degree. It is more important to be approximately right, than it is to be precisely wrong—and the numbers, whatever they are, will be wrong to some degree or other.

Explore product recommendations with in-depth demos

The product demonstrations should be done ONLY AFTER the T, I and OE estimates have been worked out above. Then, request demonstrations that provide significant proof of concept around the functionality that is intended to deliver the calculated ROI.

This approach will keep the buyer and the seller focus on what is important in order to deliver the planned ROI—and not bells and whistles that demo well but will have little or no impact on ROI. Also, the number of proofs of concept are likely to be quite small—probably five or fewer really significant things that the vendor has said they can deliver.

Such proof of concept demos may take a week or longer in configuration and execution. Avoid those things that would require heavy customization—since those should be avoided anyway. Look for the simple solutions to complex problems. If your vendor doesn’t have the innovation too recommend such simplicity, then take your business elsewhere.

As Einstein said, “Any intelligent fool can make things bigger and more complex. It takes a touch of genius—and a lot of courage—to move in the opposite direction.”

Following these simple steps, I believe, would make every ERP or other technology choice and implementation far more effective at delivering real ROI to the customer.