In a recent post, I made certain assumptions about what is or isn’t ethical in vendor-sponsored analyst research. I’d already discussed the triggering incident briefly (i.e., in Twitter direct messages and the like) with a couple of analysts I respect, namely Merv Adrian (the one most directly involved) and Ray Wang. It’s safe to say we’re in at least rough agreement.
However, a couple of comments on that post took me strongly to task. Perhaps not coincidentally, one came from a vendor, and another from somebody whose main role in the “analyst” community is to produce and publish – you guessed it! – vendor-sponsored content. One option was to just blow those comments off as nonsensical, since they weren’t really responsive to the actual post. But I think those rather surprising remarks also suggest it is time to reopen the subject of vendor-sponsored analyst research.
Vendors typically pay for white papers, webinars, podcasts, in-person speeches, etc. for some combination of five reasons:
- They want to connect with sales prospects. If Merv or Claudia Imhoff or I speak on a webinar, registration will be higher than if only company execs spoke. Similarly, you can capture more registration information from prospects who want to download a white paper if it was written by a third-party analyst.
- They want general endorsement from the analyst. If a well-regarded analyst is associated with a firm, that’s good for the firm’s image.
- They specifically want endorsement from the analyst for their marketing claims. Many of the ethical challenges with vendor-sponsored research or other content lie in this area.
- They want the analyst to do a better job of explaining something than they think they could do themselves. This is the ethically purer version of the prior point. Realistically, they often can’t be separated. E.g., most vendor-sponsored white papers will involve a combination of the two. The same could be said for webinars such as the ones I did for Aster Data last year.
- They want to give the analyst some money to enhance the relationship, and this way they get something other than advice in return. Personally, I won’t do content-creation business with a vendor unless they first buy actual consulting services (via the Monash Advantage), but I’m in the minority, and in fact didn’t always have that policy myself.
In my opinion, #1, 2, and 4 cause relatively little in the way of ethical problems. #5 is an unavoidable fact of life. But #3 raises problems that can and should be addressed head-on.
There are three forms of bias that could arise in a white paper or even a blog post, whether or not that content is paid for by a vendor. The author could:
- Say what s/he does not believe. That’s obviously a worst practice for analysts. Sadly, it is not entirely extinct.
- Omit what s/he believes. That happens all the time. If a vendor sponsors a white paper, you’d better believe that the analyst has omitted references to 1. Doubts/drawbacks and 2. Competitors. Even without payment, people to varying degrees are careful as to what they say, but the level of care goes way up when there’s money changing hands.
- Adjust tone or emphasis. If somebody’s paying you, it’s natural for your praise to be hearty and your doubts, if any, to be expressed in the most measured of tones. Again, this form of bias need not stem from immediate payment, but rather from a general desire to maintain good vendor relations. Sometimes my only quarrel with the work of analysts I otherwise admire is that they’re a little too consistently upbeat, even in their free-to-everybody blogs.
Different kinds of sponsored content tend to have different levels of bias. For example:
- White papers are commonly written with multiple cycles of vendor review. In some cases (not all), the vendor actually gets to approve or reject the paper before the writer’s payment is assured. Bias in white papers is typically rather heavy. I wrote white papers for some years, fighting off many of the pressures for bias, and it was pretty wearing, which is one reason why I’ve published only one new white paper in the last three years, and none in the last two. On the other hand, payment is commonly > $1,000/page, so I’m not going to look askance at other analysts who routinely pocket those fees.
- Lecture-style speeches and webinars can be quite independent. In the extreme case, the vendor just expects you to show up and give your standard talk from your standard slide deck. Netezza once criticized me for not slamming them ENOUGH in speeches I gave at their traveling user conference. And even when there are review cycles on the slides or whatever, those tend to be quick and gentle; about the biggest change I recall making recently under such circumstances was honoring the request “Could you please not mention our competitors by name in your PowerPoint deck?”
- Interview-style podcasts and speeches are also fairly common (especially where the analyst is the interviewer). But since I never listen to them, I can only guess what they tend to be like. (Presumably, they’re more toward the Larry King end of the interviewing spectrum than the Sam Donaldson one …)
So what would be some best ethical practices for sponsored white papers? My list starts:
- The analyst must believe everything s/he writes. Duh.
- The vendor must not have final approval over the contents of the paper. The analyst must be the final arbiter of what is said. Sure, the sponsor can decide not to publish it, and deciding that might make them less inclined to do other business with the analyst in the future. But there should be no immediate financial penalty to an analyst who refuses to say what the vendor wants.
- Sponsorship should be prominently disclosed at the front of the paper. Website disclosure is insufficient. White paper readers may get their copies from vendor salespeople or from other colleagues, and have no way of knowing the paper is sponsored if the paper itself doesn’t say so. And since people’s approach to long documents is often to check out the beginning and then – maybe – skim or sample the rest, any disclosure must be big and up front to be sure of coming across.
- The vendor should at no time state or imply that the white paper is as independent as an unsponsored one. Frankly, that’s a stretch goal, and will remain such for as long as salespeople get commissions. Even so, there’s no excuse for Oracle-level deceptive practices in this regard.
Some day, I’ll probably write another sponsored white paper. When I do, you can be assured that I’ll abide by the first three precepts, and pressure the sponsor to abide by the fourth. I’m confident that the more ethical of my analyst compatriots will do the same. I wish that I could be more optimistic that they all will.