Fun with procurement

Last week the ZMan changed tack a little and wrote about procurement in large organisations. It’s worth a read, not least because the comments did not immediately fill up with morons explaining every bad thing which occurred since the Thirty Years War is the fault of (((the Rothschilds))).

For those unfamiliar with the RFP, which is sometimes called a request for quote or even a request for information, it is a document companies produce when they wish to buy a capital product or service. In theory, the document describes the item or service, the conditions that have to be met in order to be considered and the process by which the company intends to evaluate potential vendors. These are popular in government and large corporate environments.

This struck a nerve:

If an organization or government is buying a well defined product or a commodity item, it makes sense, but for something like a complex service, then it is a recipe for failure. Even in the case of well-defined item like a machine tool, I’ve seen RFP’s that appear to be written by enemies of the issuing company. The people creating the document use it to impress their boss, rather than make a sound purchase.

The other thing that always turns up in RFP’s is the underlying assumption that the person who wrote the thing is a genius. The specifications will be hilariously narrow, which results in the request being for an exact copy of what they have now, but newer. My suspicion has been that there is a correlation between the level of specificity and the lack of understanding of the problem to be solved by the purchase. Smart companies buy products and services to solve problems. Stupid companies tick boxes on forms.

When I was in Nigeria, the acquisition form an engineer would complete in order to buy something had space for no less than nine approving signatures. Whole weeks would pass as these documents moved at the pace of a snail from one desk to another. Occasionally we’d get a question or two but they were never sensible, more along the lines of “Do we really need this?” or “Can we use two 100lb flanges instead of one 200lb?” The people reviewing the acquisition and (eventually) applying their signature added no value whatsoever; they were certainly not to be held responsible for any errors therein. But their involvement was important nonetheless, just not in the way you’d hope: it justified their existence in the organisation.

A similar thing happens with RFPs (and many other documents produce in large organisations). Thirty-three departments will all insist on being involved, requiring an endless series of meetings where each person sticks his oar in. You’ll notice this when you design by committee: every department needs to be seen to contribute something, even if it’s completely stupid or irrelevant. If they don’t, they worry they’ll not be seen as important. The result is a jumbled mess of narrow interests, pet projects, hypotheticals, and competing priorities written up by a junior employee for whom English is very much a second language. I was once involved with an integrity inspection and risk assessment project in the Middle East and tacked on the end of the RFP was a single paragraph saying the contractor had to create an entire IT system which allowed each document to be uploaded, accessible to everyone, and modifiable with automatic revision upgrades. It was obviously the bright idea someone came up with at the end of a meeting; that risk and safety consultants probably aren’t the best people to be setting up IT systems didn’t occur to any of the geniuses in Contracts & Procurement. Another requirement you see is for the contractor to train client personnel in some area, giving no guidance as to how many people and to what level you must train them.

The RFP that spawned this post was obviously the result of some serious business problem the company needs to solve. The trouble is the RFP so thoroughly obscures it, no vendor will be able to identify the problem, so they will not be able to solve it.

The arrogance of a modern company is such that they believe vendors are both stupid and liars, and they don’t need to know what the actual problem is. All they need to do is read the RFP, submit the lowest price, and be prepared to do exactly as the client tells them.

This is a good anecdote, too:

A story I’m fond of telling is about going to the initial RFP meeting for a government contract. I was a young guy and still a little green. They handed out the RFP’s and discussed the schedule. An old guy sitting next to me thumbed through the document and found the poison pill in about ten minutes. He stood up, told everyone to look at the specific section. In a few minutes everyone left the room other than me and one other guy. He was the predetermined winner. It was a good lesson.

A lot of times when a company issues an RFP it’s simply for compliance reasons. Most companies have to get quotes from a minimum three (and sometimes five) bidders, even if they have an incumbent who’s CEO is good pals with the client’s MD and they have the job in the bag. One of the first things a contractor or vendor needs to do when they receive an RFP is work out whether they’re just making up numbers on a bid-list. If you’ve never done business with this outfit before and you get a call from someone in bad English expressing disappointment you’ve not submitted a tender and offering you more time, you know the job’s gone to someone else but they need to make it look kosher. I even ran an experiment on this once, and used to turn in bids at ever decreasing prices just to see what would happen. Then I stopped responding and I got a call from a chap in contracts who asked why I hadn’t submitted a bit.

“Because it’s obvious we’re just making up numbers on a bid-list while every job goes to that company you always use,” I said.

He was most indignant.

Liked it? Take a second to support Tim Newman on Patreon!
Share

24 thoughts on “Fun with procurement

  1. Basic rule in IT as a vendor is that if you (or someone who resells/understands your product) did not help write the RFP you won’t win the contest so there’s no point in trying.

    I spent a lot of time being told to respond to RFPs that were written by Cisco/for Cisco equipment when I worked for Cisco competitors. It was utterly pointless writing a response because even though our product was almost certainly better in any test of price/performance or even (in some cases) basic functionality, you would never win because there was always some trick that could only be solved by Cisco. Said trick was usually irrelevant to the actual task (e.g. asking for support of a routing protocol that wasn’t on the customer network and never would be) but was marked as something that was mandatory.

    I did one time come up with a solution for the trick the response was to modify the RFP so my solution no longer worked.

    We did occasionally win these biased RFPs but the way we won them was much the way that Cisco sometimes won the ones we wrote – namely taking the CFO, CTO etc. out for an expensive jolly such as the Monaco Grand Prix

  2. Procurement is a honeypot. The stuff you can steal is brand new. Hence, perhaps, the nine signatures. Difficult to organise the division of the spoils among so many.

  3. The RFP that spawned this post was obviously the result of some serious business problem the company needs to solve. The trouble is the RFP so thoroughly obscures it, no vendor will be able to identify the problem, so they will not be able to solve it.

    Sometimes (see above) this is deliberate. They don’t want other solutions to the problem because they’ve already made up their mind. They just want bids from other vendors so they can claim the fix wasn’t in.

    One thing I have found is that if the need is urgent and/or a critical part of the organization’s main business then the RFP tends to be fairer and less convoluted. In those cases you’ll typically be told relevant facts and be judged on metrics that matter. If the RFP is a mess that usually means it’s being produced on behalf of otherwise powerless drones who now have a chance to make others jump through hoops for them as in this Dilbert – http://dilbert.com/strip/1996-06-06

  4. I spent a lot of time being told to respond to RFPs that were written by Cisco/for Cisco equipment when I worked for Cisco competitors.

    Ah, but this doesn’t always work out the best for Cisco. I’ve been invited to help write a tender (for free) on the assurance that this would put us best-placed to win it; we were invited because the client has no idea how to write the technical specification or methodology. The tender then goes out to Ali Bin Flip-Flop who says “Yeah, I can do all that for half what Tim’s mob charges” and gets the job. That he can’t deliver doesn’t matter much to the procurement people, who have “saved” the company 50%. It’s the project people who will have to deal with Mr Flip-Flop, and who cares about them?

  5. Reading all these tales of middle management in large corporations, one does get the impression that most of these people are not adult enough to be middle management in large corporations. Quite depressing. But can’t see it lasting. Either the corporations will be out-competed by smaller, leaner, more efficient competitors or….
    It does occur that this is a whole strata of employment that going to be replaced when strong AI comes of age. Taken individually, none of the decisions these people make are complex. It’s the system as a whole is complex. Personally, I wouldn’t give it 5 years.

  6. It does occur that this is a whole strata of employment that going to be replaced when strong AI comes of age.

    That’s where you’re wrong: these jobs are already unnecessary, hence these people are already part of a giant job creation program put in place because there’s nothing else for them to do. The more robots take over, the more humans will be employed in “compliance” roles.

  7. BiG’s rules of cost proposals (scope may be expanded at any time)

    1: The clearer the client’s specification of what they want the further it is from what they actually need.

    2: The client’s certainty as to what they want is inversely proportional to their certainty about what it will cost.

    3: The client will always fall for your competitor pulling the old trick of deliberately underestimating the initial costs then requiring 12 change orders.

    4: The amount of billable work to be done is inversely proportional to the effort required to secure the work.

  8. 5. Rule 3 applies in particular when long-term “strategic partnerships” are being dished out. The cost of your competitor’s change orders doubles once the client has fired their entire internal team because the strategic partner looked cheaper at the bid stage.

  9. 6. Buying your client drinks at a conference is corruption. Your competitor buying your client drinks at a conference is networking.

  10. “34: ”

    I’m tempted to say “challenge accepted”, but I think our genial host is more likely to know where to find the relevant examples.

  11. I did actually mean the whole thing, Tim.
    At one end of procurement you have a guy who needs something. At the other end, you have a guy who can supply what the first guy needs. The rest of it is administration. Matching the appropriate supply to demand, pricing etc would be right up an AI’s street. It’s not as if we don’t already have an efficient analog computer that does this sort of thing. It’s called a market.

  12. I think our genial host is more likely to know where to find the relevant examples.

    And to think, I’ve spent a fortune on my PR campaigns!

  13. At one end of procurement you have a guy who needs something. At the other end, you have a guy who can supply what the first guy needs. The rest of it is administration.

    Ah yeah, fair enough. It does get complicated with big bits of equipment and services, but most of the people who get involved have no stake in the outcome.

  14. Usually procurement’s only care is to get the lowest possible number at the bottom of their spreadsheet. There are many tricks to achieving this, even on the client side, such as assigning work packages with high granularity, thus dividing a complex project up among more suppliers than is good to get the job done. This ends up as a mix of externalising created costs to those suppliers who have to now talk to each other, and additional billing.

    Whether the actual end cost has anything to do with the initial agreed cost should be the metric used to assess the effectiveness of procurement.

  15. Between this and political donations, it seems like the low-corruption west has actually just put in a lot of effort to hiding corruption, but it’s all still there. Maybe we’d be better off going back to some kind of open patronage system. What would be the difference?

  16. Some larger organisations do run decent RFQ (procurement) and Invitation to Tender (ITT) systems for onsite subcontractors. In previous lives I used to do a considerable amount of surveillance of vendors and knowing full well what is in every RFQ, the only variable being the scope, the drawing, the spec and the delivery date it was a great way of ensuring that, that is what was sent to site and keeping the vendor in check. By keeping them in check I mean walking through the RFQ with them, ensuring they understood it and quite often helping them interpret parts of it or ringing up our guys and resolving queries or waiving parts of it. If you are sitting in a workshop in South Korea with a bunch of fabricators that are not ofay with the finer points of your RFQ they really do appreciate you helping them out with it and the dinners get better as well.

    Same for onsite subcontractors (ITT) you could generally manage them on site without even having read their Subcontract because you knew what was in it. Say you are doing a site inspection with an irate subcontractors’ manager and he unexpectedly complains to you about something you should be able to start conditioning him right from the start with your initial response because you know in general what his engagement terms are, issues can then be quickly nipped in the bud or escalated. If an issue has legs of course you would refer to it but like RFQ’s the only part that should change is the scope and schedule, the other 80% i.e. terms and conditions, progress payments, document submittal and approval, quality, safety, environmental, IR, cultural heritage, community relations etc are standard per project. This is typically where the problem lies not in the scope of work which the vendor or subcontractor is normally proficient in delivering.

  17. “What would be the difference?”

    The biggest issue that we have in the west as opposed to cultures that have had a system of commission for millions of years that is perfectly acceptable and not frowned upon is accounting for the payments on the books. Hands on fees, commissions etc can be shown as line items on tenders, budgets, cash flows and cost to complete reports in say places with Javanese or Arabic cultures, in the west you must be far more creative in accounting for your Aston Martin, giving and receiving of gifts, donations etc. It’s a big problem these days.

    I am in a mid-sized international company and our accounting standards vary considerably by region. We had a good laugh once when one of our non-Australian directors quite openly said during a corporate governance awareness session held inside our lawyers office in Brisbane that if he is in Mozambique and some armed general wants something from him, he is getting it there and then.

    Plus, it is an established fact of doing business that many larger singular projects in non-OECD nations must strike incentivization deals directly with the local port customs workforce and also that the initial deal isn’t the final one, this works when you can show a contingency allowance for escalation duty in your budget, it doesn’t when you can’t. We never, ever take on customs risk in these countries and we always pass this risk on to our clients at the time of tender, its a walk away if we can’t.

  18. Buying expensive stuff needs controls and a consistent process to ensure the requirement is satisfied at an acceptable price without incurring risk of jail time, fines or disciplinary outcomes.

    There are several ways to achieve this. RFx seems to be the least bad for most large organisations.

    The problems arise not because the process or policies tend to be daft but the people following them don’t engage their brain to constantly remind themselves and everyone what the shared desired outcomes should be.

    As with most corporate standards, people have a tendency to turn into “airport zombies” in response to procurement policies and process; by which I mean that semi-conscious state you enter once you’ve checked in and dropped your luggage, wandering between checkpoints and responding to each demand from the official in charge until you’ve reached your destination.

  19. William, the attitude often remains after procurement.. It’s amazing how many organisations are satisfied with utter shit for an outcome, provided there were no deviations from the SOP.

  20. The problems arise not because the process or policies tend to be daft but the people following them don’t engage their brain to constantly remind themselves and everyone what the shared desired outcomes should be.

    This. I work in IT and I periodically shock people by recommending we blatantly violate security best practice. I point out that the ultimate goal of every company is to make money, and if complying with all the security crap is provably going to cost us much more than taking the risk (including things like a reasonable evaluation of the risk of a penetration and/or failing a mandatory security audit), then take the risk.

Comments are closed.