Some thoughts on Gartner

I sometimes flatter myself by thinking this blog offers interesting insights when in fact I know a key reason that many of you subscribe is to save time:

You could spend dozens of hours talking to management and combing through annual reports, proxies, and transcripts, or I can just do it and share what I’ve learned. If the average scuttleblurb write-up requires 50 hours of work and I do 24 of them a year, then each of you save 1,200 hours of time for $210. Even if only half my posts resonate, that’s still a pretty good deal. This arrangement works for me too because I can sell the same write-up an unlimited number of times.

Gartner shares some similar features. The company decides which topics across IT, logistics, HR, marketing, sales, and finance to explore based on client feedback, session attendance, popularity of search terms on its site, etc. and has its analysts gather intelligence on product/market landscape and cobble together digestible reports that can be sold countless times. Gartner’s $200k average annual contract value[1] is inexpensive relative to the $50mn+ IT budget of most of its clients, who reference the company’s research before making high stakes decisions. 

So a Chief Security Officer might peruse some Gartner research and arrange a half hour call with one of Gartner’s security experts before explaining Zero Trust to her Board. Or the Chief Technology Officer of a large insurance carrier might consult Gartner as she deliberates on a disruptive and consequential undertaking whose failure could cost her job, like whether to adopt a modern core processor and which one. The IT staff may deride Gartner as a vendor PR, part of the vacuous dog-and-pony spectacle stipulated by the suits…after all, it’s not like Gartner is hands-on-the-wheel installing the firewalls or running the container instances on different clouds and then tailoring a set of recommendations. But this is beside the point since an executive is looking for more than technical feedback…he wants customer feedback and a detailed understanding of the state of the market, including each vendor’s product roadmap and ecosystem support, details that may be elusive to someone narrowly focused on specs.

Moreover, myopically focusing on the technical merits of a Gartner opinion also ignores another important job to be done. A corporate executive might pay for Gartner’s advice to cover his ass on a mission critical IT decision, much as a CEO might pay Goldman Sachs for a fairness opinion on an M&A transaction he already knows he’s going to do. In the same way that Moody’s ratings have become a standard way of assessing relative credit risk, the Gartner Magic Quadrant (GMQ) has become a succinct standardized way of mapping the relative strengths of different vendors (public cloud providers, HCM, ERP, core processors, No-SQL databases, etc.), which comes with a nice reflexive advantage: the more CIOs who rely on Gartner to validate their decisions, the more likely Gartner is to get hired by the next CIO, further entrenching the company’s reputation as the “go to” source for informing (or even justifying) IT decisions. 

So executives depend on Gartner for both the objective quality of its research and the social proof/downside protection it affords. Like, I’m sure IT buyers and managers are genuinely seeking product/vendor guidance and want a trusted third party to help them sort things out and whatever, but they also want to minimize career risk while doing so (you are less likely to be fired for failing conventionally). These two things aren’t as neatly segregated as I’ve described – if Gartner starts publishing shitty work and consistently gets things wrong, the brand reputation it has cultivated over the last 40 years will deteriorate – but it’s almost certainly true that no CIO of a F2000 company would give scuttleblurb (scuttleblurb who?) a second look, even if I produced evaluative IT research superior to Gartner’s.

But this can’t be the whole story. Forrester and CEB have also developed well-established brand names over decades and yet neither has created nearly as much enterprise value as Gartner. 

So then what about operational excellence? Business analysts usually draw a sharp distinction between operational excellence and competitive advantage because the former is something that can, in theory, be replicated by any organization willing to change the way it does things. But we know in the real world that the reason two companies in the same industry with access to the same moats often deliver radically different results often comes down to differences in organizational structure and behavior. Rather than dismiss this variance as something that is supposed to be competed away, perhaps we should accept operational excellence as a symptom of culture…and sometimes differences in culture give rise to organizational postures that then help create enduring advantages in industries where you might least expect…for instance, Old Dominion, Fastenal, and Ryanair’s culture of cost discipline accelerates local scale economies that allow each company to trounce its less disciplined rivals in logistics, distribution, and airlines, respectively.

Gartner is an example of this. Its reputation as a trusted advisor for enterprise executives is sustained by something distinctly unmoaty: a durable sales apparatus that for whatever reason its competitors seem unable to replicate. I am not referring to Gartner’s salespeople, 20%-25% of whom churn off every year (vs. ~5% for research analysts), but rather the system of recruitment and training, the best practices sales playbook that every sales rep imbibes during their first 6 to 8 weeks of training. 

[Fwiw, a Twitter friend recently shared an illuminating anecdote with me. A friend of his works as a salesperson at Gartner. During her 2 months of training, she and a class of fellow recruits were isolated in a Florida hotel, where they were split into teams that competed against one another in intense, high pressure sales games every day (“people were crying”). Around 10% of her class was fired within the first 2 weeks, but the others who made it through bonded over the shared trauma and became part of an elite cohesive sales organization whose members are highly coveted by other firms.]

Sales is Gartner’s overriding priority for good reason. Some enterprise products, like an ERP system or a firewall, are taken for granted as necessary, to be RFP’ed and evaluated in the normal course of business. Consulting engagements are more ad hoc but similar to most other enterprise products in that by the time you think to call McKinsey or BCG, there is little confusion about the problem you are trying to solve. 

Research subscriptions are

different. It’s not like C-level execs

start the year thinking they need research explaining the state of the

distributed database market, or expert advice on building resilient supply

chains or inculcating best practices for new hire training. They aren’t sifting through a stack of

Gartner reports and meticulously evaluating the quality of that research to

Forrester’s. No. While Gartner technically competes with

Forrester, a sluggish rival with less than a tenth of Gartner’s revenue, most

of the time it is up against non-consumption because research is the kind of service

that most enterprises don’t know they want or need. It requires a skilled and aggressive

salesperson to get the attention of a C-level decisionmaker, understand his priorities,

and craft a compelling narrative that convinces him to keep abreast of certain topics

on a recurring basis. Gartner has refined

its sales process over many decades, leveraging it to push research from acquired

companies like AMR Research and Burton, both of whom struggled to scale as independent

companies.

I don’t want to overplay the culture card. But Gartner consistently produces 100%+ dollar retention rates in an industry that doesn’t enjoy inherent switching costs; pushes through price increases; and generates far more revenue and profit growth than its peer group leads me to suspect that there is something different about the way it does things. To the extent that’s true, then the risks to Gartner’s business value comes down to dramatic changes in culture or in the value of third-party IT research. Re: the former, just the opposite appears to be true. With technology infusing every facet of the enterprise, sr. execs are evaluating more technology than ever as they effect radical changes to their IT architecture to remain competitive (cloud! data! IoT!). 

Prior to 2017, you may have been inclined to dismiss the latter concern too – Gartner’s forays into supply chain and marketing expertise have worked out (and in any case were too small to really matter) while its consulting and conferences businesses are sensible ways to monetize and reinforce the core research business….but then Gartner went ahead and bought Corporate Executive Board (CEB) for $3.5bn in its largest deal to date.

CEB differs from Gartner – which generates expert analysis on IT, supply chain, and marketing and sells that research per seat – in that it produces case studies and best practices recommendations for HR, Sales, and Finance departments and sells on an enterprise-wide basis. At the time Gartner acquired it, CEB was a withering company with decelerating revenue[2] and contracting margins. Ironically, even as CEB sold best practices on HR and sales, it struggled to optimize its own sales force. The company’s relationship managers were juggling basic administrative search tasks rather than establishing new relationships while its front-line salesforce lacked adequate training. Moreover, the company made a bunch of dumb acquisitions that led to years of impairment charges, mistakes that I imagine impugned the company’s product credibility (would you take investment advice from an advisor who blew up his fund?). So besides the standard cross-selling and cost synergies, one of the key sources of upside from this acquisition was the application of Gartner’s best-in-class sales and operational playbook to CEB’s product suite.

But the problem is that research is

not fungible across sales teams. In

2012, Forrester tried re-orienting its sales process around account rather than

function – so rather than having a dedicated IT person sell IT products to

companies A, B, C, and D, it had one salesperson selling IT, marketing, etc.

into account A; another sales person selling those same disparate products into

B, etc. Gartner also tried doing this in

2010 after purchasing AMR Research – it had its IT sales reps sell supply chain

research to heads of manufacturing and logistics – which turned out to be a

fiasco. This means that Gartner can’t just

have its polished IT, marketing, and supply chain reps push CEB’s HR, sales,

and finance research. Instead, it needs

to indoctrinate CEB with Gartner best practices, which is very heavy lift. It takes 3 years for a Gartner rep to become

fully productive, implying that CEB’s salesforce won’t be ramped until 2021. 

Still, there are several reasons to be optimistic. 

First, while IT research is the

dominant value driver, accounting for over 80% of the Research segment’s

contract value and even more of its profits, Gartner acquired its way into supply

chain in 2002, organically built digital marketing expertise in 2012, and

succeeded in both. As of a few years ago

(the last disclosure we have), supply chain and marketing were growing contract

value by more than 25% at economics comparable to Gartner’s core IT

business. 

Second, the company has been letting CEB’s

legacy lower-margin Marketing products churn off and replacing them with higher-priced/better

retaining “GxL” subscriptions (i.e Gartner for HR Leaders; Gartner for Finance

Leaders; Gartner for Marketing Leaders). 

The latter basically consist of research and advisory services specially

tailored to division heads….so via GxL, Gartner would tailor existing IT

content to teach a CFO about financial accounting modules of different ERP

systems or something, or leverage CEB’s HR best practices research to help a

CTO nurture talent.

For the last few years, contraction in the legacy business has more than offset growth from GxL, giving rise to meager low-single digit contract value growth in the GBS segment (Gartner Business Solutions…non-IT research). But sometime this year, GxL will account for more than half of GBS contract value.

The mix shift toward better retaining

GxL business is also fueling stronger GBS wallet retention[3], which has

climbed from 90% in 2016 to 101% in 2019, and salesperson productivity, which

spiked from $14k in 2018 to $67k in 2019. 

I would expect that as GxL comes to dominate GBS over the next 2-3

years, Research revenue and CV growth will trend towards low-teens growth.

Third, after frontloading growth in the GBS salesforce, Gartner is growing headcount by just 8% this year while bringing GTS sales rep growth to just 8%-9% vs. ~13% over the last 7 years. Also, under its new “just-in-time” training program, the company will be getting new hires out in in the field 4 to 6 weeks earlier than before, the equivalent of adding 3%-5% to headcount[4]. What this means is that sales rep productivity and contribution margins should improve as the seasoning of previous years’ hires are offset to a far lesser degree than before by relatively unproductive new hires. 

Now, given the labor-intensive nature of this business, there are limits to profitability gains. Gartner Research’s contribution margins[5] of ~70% (meaning the company realizes ~70c in variable profits for every dollar of incremental revenue) are unlikely to inch much higher because a large chunk of incremental revenue must be reinvested in client service to sustain the 6 year life of an average customer. But even so, I would expect that boosting salesforce productivity and closing the sizeable productivity gap in GBS – where contract value per sales rep is just 2/3 what it is in GTS – should nudge Gartner’s EBITDA margins from 15% today to their pre-CEB levels of 17%, with EBITDA dollars growing a point or so faster than revenue on a normalized basis.

According to management, there are 100k enterprises with IT budgets greater than $10mn. Gartner sells to just 13k of them. In dollar terms, the market for core IT research is $55bn (which management derives by estimating all the subscriptions that could be sold to enterprise CIOs and their direct reports), many times larger than the $2.6bn of IT contract value that Gartner is currently realizing. The market opportunity is even more compelling for GBS, which has a base contract value of less than $700mn despite covering a broad range of domains, including supply chain, marketing, sales, and finance. 

Gartner’s Conferences and Consulting

businesses draw their strength from Research, but also reinforce the Gartner

brand and drive deal flow. 

The Conferences business has done well over the last 8 years…

….Consulting[6] less so…

Both businesses are economically sensitive. During the last recession, their revenues and gross profits suffered significant declines:

Conferences saw its revenue drop by

40% and its contribution profits cut in half from 2007 to 2009. But the segment regained previous highs

within 2 years. Meanwhile, Consulting

revenues fell by just 15% from its 2008 peak, but it took 6 years to return to

peak…and its contribution profits are still below 2008 levels, as consultant

utilization and revenue per billable consultant have steadily declined, suggesting

structural problems in this business.

So while Consulting strikes me as an

iffy business, I think Conferences will continue to be a fine complement to Research,

as people love getting together with peers who share common interests

(Berkshire Hathaway Annual Meeting, Manual of Ideas, Grant’s). For modeling purposes, I assume Gartner’s

Conferences revenue plummets 70% over the next year due to COVID and that it

takes 4 years for revenue and contribution profits to regain their 2019 peak (vs.

2 years during the last recession). In Consulting,

I assume revenues decline by 40% and contribution profits get cut in half over

the next year and that neither recovers 2019 levels over the next 4 years.

COVID’s impact on Research should be relatively tame. Gartner’s Research segment has experienced just 2 down years over the last 2 decades: in 2002, when revenue fell 7% as enterprise IT spending retrenched at a time when the company was still delivering its products through print and CD-ROM, and in 2009, when revenue fell 4% (1% in constant currency) in the throes of a severe credit contraction. 

[Research contract value is a leading

indicator of revenue and is defined as: “the value attributable to all of our

subscription-related research products that recognize revenue on a ratable

basis. Contract value is calculated as the annualized value of all subscription

research contracts in effect at a specific point in time, without regard to the

duration of the contract”] 

Moreover, even as Research revenue declined by 4%, contribution profits increased by 1% as the company pushed through price increases and managed its costs[7]. In 2010, revenue rebounded 9% on an organic basis, achieving new highs (even excluding contribution from the AMR and Burton acquisitions). And so I’m going to assume Research revenue droops by 15% over the next year (worse than 2002 and 2009) before regaining its highs 2 years later, with gross margins inflating by ~1pt as management whittles down costs as it did in previous downturns. 

Adding contribution profits across segments and assuming SG&A declines at 60% the rate of revenue in year 1 before growing at 70% of revenue thereafter, I can get to 16% EBITDA margins by year 5 compared to 15% margins in 2019 (which includes CEB in turnaround mode) and 18% margins in the pre-CEB 2017 peak. This gets us to $5.60 in per share earnings (ex. amortization of intangibles), which at 30x + accumulated cash implies a “meh” 9% return off the current price ($114) (in this rate environment, 30x earnings feels about right for a moaty company with nominal reinvestment requirements generating low-teens organic EBITDA growth). I have no issues with Gartner’s liquidity profile. Trough year EBITDA of $350mn still covers interest expense by 3.5x and there are no meaningful debt maturities until 2022. Cash flows plus the $1bn untapped credit facility should comfortably absorb any of the $980mn of negative working capital that unwinds over the next year. 

Forrester trades at a discounted trailing multiple vs. Gartner (16x 2019 earnings vs. 27x), but it deserves to. In 2007, as the much smaller player in independent IT research, the company tried to differentiate itself by catering its research to particular roles within an organization rather than focusing on the topic alone (“here’s how the Chief Investment Officer should think about Information Security” vs. just “Information Security”). Forrester also offers a few other things that Gartner doesn’t, like you can pay two thousand bucks to take Forrester’s Customer Experience, Zero Trust security, etc. course and earn certifications that can be pinned to your LinkedIn profile. Or you can leverage Forrester’s comprehensive survey data to better understand your customers or which technology purchases business leaders are making in different parts of the world. But without a robust sales apparatus to disseminate its products, none of the company’s differentiating selling points has made a difference in closing the competitive gap.

Whereas Gartner is a vibrant enterprise, reinvesting into profitable revenue growth, Forrester is an inert entity, recycling all its cash flow into buybacks and dividends. Whereas Gartner has grown its sales headcount by ~13%/year over the last 7 years, roughly matching its organic revenue growth, Forrester’s salesforce has grown by just ~2.5%/year. From year-end 2011 to 2018, a period that included no acquisitions, EBITDA declined from $50mn to $35mn, even as revenue compounded by ~3%, from $283mn to $358mn. Per share earnings went nowhere in the decade from 2008 to 2018[8] while Gartner’s have nearly quadrupled over the same period. 

A lot of this underperformance ties back to some ill-conceived moves in 2011, when Forrester made it harder for salespeople to earn commissions and re-oriented the sales process to a hybrid model (as discussed earlier). These missteps were supposedly reversed in 2013/2014 and yet profitability has continued to stagnate:

And SiriusDecisions, a marketing/sales research business that caters to the tech and manufacturing verticals and which represents Forrester’s largest acquisition to date, has performed worse than expected, its ~$83mn of revenue contribution in 2019 falling short of ~$100mn contribution that Forrester’s management expected at the time of the deal’s announcement in November 2018, as the salesforce has churned off at a higher than expected pace. Sirius in fact generated less revenue in its first full year under Forrester than it did during its last year as a standalone company. While CEB has been sort of a disaster, Gartner at least has sensible and aggressive plans to turn things around whereas Forrester appears to just be hoping for the best.

Footnotes

[1] ~$20k/year gets you read-only

access to research; $150k gets you a dedicated analyst who may have once served

as a CIO at some point in their career

[2] its ~10% organic growth flattening to low-single digits in the 2 years prior to being acquired

[3] Gartner defines wallet retention as: “a measure of the amount of contract value we have retained with clients over a twelve-month period. Wallet retention is calculated on a percentage basis by dividing the contract value of our current clients, who were also clients a year ago, by the total contract value from a year ago, excluding the impact of foreign currency exchange. When wallet retention exceeds client retention, it is an indication of retention of higher-spending clients, or increased spending by retained clients, or both. Wallet retention is calculated at an enterprise level, which represents a single company or customer” (2019 10-K)

[4] new hires typically go through 6

to 8 weeks training before they start selling. 

Under this new arrangement, sales reps are put in the field after 2

weeks of upfront training, the remaining 4 to 6 weeks spread out as 1 day/week

sessions.

[5] Gartner defines contribution

profits as “operating income excluding certain COS and SG&A expenses,

depreciation, acquisition and integration charges, amortization of intangibles,

and Other charges”. The sum of

contribution profits across segments very closely approximates the company’s total

gross profits over time.

[6] typically $300k-$500k 7-to-9 month

gigs that deal with IT sourcing and architecture strategy

[7] nearly all of Gartner’s costs are

in the form of semi-variable labor, which can be flexed to some degree in the

event of a major top-line hit. In 2009,

Gartner partly offset its 11% total revenue decline with a 10% reduction in

total costs

[8] I used 2018 as the end year because 2019 includes Forrester’s $238mn acquisition of SiriusDecisions

Disclosure: At the time this report was posted, Forage Capital did not own shares of Gartner or Forrester. This may have changed at any time since.