How Bad Can It Be?

From the June-July 2015 issue, page 34

Several years ago I was in Victoria, B.C. doing a presentation to the local chapter of the Canadian Association of Management Consultants about selling professional services. As usual, part of my discussion included bashing the costs, ethical issues and inaccuracy of the Request for Proposal (RFP) process.
As we were wrapping up, one of the attendees raised his hand and said, “Cal, I would bet money that if you added up the cost of all the time spent writing proposals by all the proponents to an RFP, then those costs would be greater than the fees available in that RFP!”
That thought hit me like a Dustin Byfuglien slap shot and tormented me for years because I knew he was right. But how could I take this idea to the world?
Finally one day I had the answer.
I designed a simple experiment to try and benchmark for professional services firms how frequently the total costs of writing a proposal exceed the fees for the winner.
For example, if 10 firms spend $5,000 each writing proposals in response to an RFP, the total proposal writing cost is $50,000. If the fees available to the winner of that RFP are $40,000, then the proposal writing cost (what I call the “industry response cost”) is greater than the fees to the winner. I consider that to be an indicator of a very low-efficiency process. What I wanted to see was just how frequently this inefficiency happens.
So I analyzed the RFPs in three professional services categories available on MERX (the electronic tendering service for those bidding on Canadian government contracts) on one single day. I analyzed for three types of service: Professional, Administrative and Management Support Services (PAMSS); Information Processing and Related Telecommunications Services (IT & Telecom); and Architect and Engineering Services (A&E).
I reviewed hundreds of RFPs and did an estimate of the hours I believed it would take to respond to each RFP. I multiplied that by a standard hourly rate to come up with the cost of writing a typical proposal for each specific RFP. I then multiplied that cost by 50% of the total number of individuals that downloaded the RFP. In other words, I assumed that only half of the companies that downloaded the RFP would actually submit a proposal) which gave me the industry response cost for that RFP.
With that I was able to compare the industry response cost to the fees that the winning firm would receive based on a budget stated in the RFP. If no budget was stated then I did not use that RFP.
The results were very interesting, as shown below.
Outside the obvious observation — that the frequency indicated in the table above is way too high in all categories (67% for A&E!)  — there exists another interesting question: Why are the numbers so different? To that I do not have a good answer but I do have two thoughts.
First, the numbers seem to increase as the complexity of the professional services increases.
While architecture and engineering consulting can be a very complex process, when you layer on technology variables (hardware, legacy systems, software, networks, security, etc.) or design and construction variables (materials, location, weather, building codes, legislation, etc.), it’s easy to see how the proposal writing process could take much longer for A&E firms compared to the companies in the other two categories. (As well, I am sorry but I must say that architects are notorious for giving away too much thinking up front, especially in things like design competitions, which other sectors like advertising have somewhat successfully tried to outlaw for years.)
Second, IT and A&E projects are often bigger budget projects, so it makes sense that their proposals would need bigger budget as well. As  scale increases, and variables increase, so too might proposal writing cost volatility (possibly increasing exponentially instead of in a linear fashion when complexity increases).
The real purpose of these numbers, however, is not to nail down what the cost actually is, but simply to accelerate the discussion about how we can reduce the industry’s response cost for professional services projects and ultimately increase the quality, integrity and value to the end user (which is often the taxpayer).
I am not suggesting we just award projects without some sort of process, but I am suggesting that we make the process better.
And I believe that a qualifications-based selection (QBS) process, with a budget disclosed to the vendors, and a price negotiated with only the most qualified vendor, will be a significant improvement compared to the current use of a priced-based RFP in the professional services buying process.
Why? For three simple reasons.
(1) Lower cost and better quality decision-making. It takes a lot of time and money to try and predict the price of a project, and research by the American Public Works Association (APWA) and American Council of Engineering Companies (ACEC-US) has clearly indicated that price is a poor selection criterion. So by removing price as an evaluation criterion we remove a lot of the proposal writing cost while also improving the quality of the selection outcome.
(2) Better fraud protection. Price is easy to manipulate, whereas expertise is easily and objectively evaluated (with a trained eye) and also difficult to fake. So selecting based upon qualifications and not price, provides much better protection against procurement fraud  — protecting the buyers from allegations of impropriety and the end user (again, often the taxpayer) from the misuse of their funds.
(3) Easy to implement. Changing to QBS is almost a “costless change” as it requires no new people, no new technology, no new legislation, and no new policy writing. Just a decision to change, some support from management (and maybe politicians), and a bit of training on how to shift from a Request for Proposal, to a Request for Qualifications using a QBS-based scoring rubric.
So what are we waiting for?      cce

Cal Harrison, MBA, CMC, is the president of Beyond Referrals, a Winnipeg based consulting firm. Cal@BeyondReferrals.com