Author: Michael Feldstein

I did not want to write this post. I really didn’t.

For starters, I’ve been delighted to get away from this sort of corporate analysis and accountability writing, knowing that Phil has it well covered on his blog. That goes doubly for LMS inside baseball. Second, having recently announced a major sponsorship from one of Instructure’s competitors, there’s no way I can write about this topic without the appearance of a conflict of interest, particularly when my assessment is negative. Third, After watching the fights about what the acquisition means breaking out on social media, I have about as much desire to insert myself in the middle of that as I do to stick my index finger into my garbage disposal.

But people keep asking me to write about Instructure’s prospective acquisition by Private Equity (PE) company Thoma Bravo. And it’s a consequential moment. One of only a very few mainstream LMS providers is at an inflection point. Perhaps more than customers realize. I have become increasingly worried that, by going through with this particular sale at this particular time, there is a very high risk of destroying the company’s value to customers and shareholders alike.

So here I go. Stickin’ my finger in the garbage disposal.

In this post, I’m going to explain why I worry the current deal on offer to take Instructure private runs a high risk of being bad for everyone.

Instructure probably needs to do this at some point

Let’s start with a basic reality: The reason that Instructure is moving toward a sale to private equity is that the company is at an inflection point and, despite its history of success, could very easily fall apart. Most Instructure customers don’t see this; what they see is a company that keeps growing and, on the surface, doesn’t seem to have changed dramatically. But major shareholders and board members have a different perspective. They see financial problems down the road, and they also see that life under private equity could buy the company the time it needs to address these challenges in ways that staying publicly-traded might not. There is some debate among shareholders about how quickly this move must be made and how to best accomplish it. But I haven’t heard objections from these stakeholders that taking the company private is a bad idea in general. I agree with them on that point. I personally believe that a sale to the right PE firm is probably a good idea in principle, both for shareholders and for customers.

Instructure’s competitors have made a big deal out of the fact that the company still isn’t profitable after all these years. That’s a bit of a red herring. Instructure has chosen to reinvest profits in the company. When you do that, the “extra” money the company has made from a sale is no longer called “profit.” But that is a choice.

The real problem is that selling Learning Management Systems is, in and of itself, not a very good business. Yes, Instructure could have been profitable sooner. But not very profitable. The way that LMS companies become financially healthy in the long run—to the extent that they do—is by selling other products and services to their existing LMS customer base. They increase the number of dollars per customer that they earn by selling those customers other things. (We’re going to return to that last sentence later in this post, because even though it is true, it is framed in a way that often leads EdTech Private Equity acquirers to make very bad decisions.)

Instructure has a strong and growing customer base (even if the rate of growth has slowed recently). That’s why investors are attracted to them. What the company lacks is a portfolio of other products and services to sell. They have a good foundation, but they need to build on it. So there are a few critical questions that I am going to address in this post which are relevant to everyone who wants to see Instructure become more valuable rather than less:

1. How did the company manage to grow so dramatically for so long?
2. What does the company need to do in order to avoid breaking the engine that enabled them to acquire and retain customers?
3. What would a credible plan for increasing Instructure’s future value to customers look like?
4. What is the gap between a credible plan and what Instructure has presented so far?
5. Why is that gap bad for current shareholders and dangerous for customers in the context of a potential acquisition?

Before I get to these questions, though, I want to address a source of anxiety that I am seeing on social media that I think is misplaced. There is a narrow and outdated notion of what PE does that is leading to some unrealistic fears among some concerned academics. While there are real risks to worry about here, it’s important to focus on the fears that are realistic.

Private equity is…equity that is private

From a definitional standpoint, a private equity (PE) company purchases equity—stock shares—of the companies it invests in, but it does so outside of the publicly traded stock markets. For many years, private equity companies tended to follow one of only a couple of well-defined playbooks. Sometimes they would be house flippers, fixing up an undervalued house and then turning it over (relatively) quickly for a profit. Sometimes they would be junk dealers, stripping a car with a broken engine for the parts that still had value. Sometimes they would be slumlords, extracting as much “rent” from the asset as they could while putting as little investment in maintenance as they could. While the word “slumlord” is (intentionally) pejorative, some of these models could work out fine for the customers and employees, depending very much on the specifics. More on that in a bit.

Now that the private equity markets have reached over three trillion dollars (and growing rapidly), there is a lot more diversity in the kinds of investors buying equity privately. Their goals, risk tolerance, and strategies vary pretty widely. There are, for example, family offices where rich people invest their money in individual companies rather than on the public markets. Depending on the goals of the family, the behavior of these funds can be quite long-term-minded and benign. Or not. You just can’t conclude much from the label “private equity” by itself anymore.

Here’s another way to think about it: Academics tend to be suspicious of venture capital, private equity, and IPOs. Basically, any funding is anxiety-provoking because of the potential strings attached. In Instructure’s case, they earned a spectacular reputation with customers while being venture-backed. They’ve done less well, but not horribly, since their IPO. What can we conclude, then, about the relative badness of VC-backed companies versus publicly traded ones?

Uh…not much. It comes down to the alignment of incentives between the investors and the customers. Understanding that will help to clarify which fears about this particular acquisition are realistic and which are not.

Instructure has no parts to fleece

One concern I’ve read is that the new private equity (PE) owners will sell off parts of the company. To which I reply, Which parts are you worried about them selling off?

Canvas?

No. There would be no company without Canvas.

Arc?

Are people massively freaked out about the potential sale of Arc? I’m guessing not.

Portfolium?

I’m not saying anything bad about the product itself (or any of Instructure’s products), but is all of this hand-wringing about the possible sale of Portfolium? Doubtful.

Bridge?

Hmm. Let’s set Bridge aside for a moment, look at a situation with an LMS company where selling off parts actually was a viable business strategy, and then return to it.

Blackboard’s PE owners sold off CashNet less than a year ago. It was an absolute nightmare for the company’s bread-and-butter education customers.

Oh, wait. No, it wasn’t.

Blackboard Learn customers, by and large, didn’t even notice that CashNet was gone. It was an asset that had more value to those customers in the form of the cash that Blackboard got for it than in the actual ownership of the product. This was a case in which selling off a part of the company was both a good strategy from a financial perspective and inconsequential at worst to customers.

In fact, Instructure is struggling now largely because it doesn’t have enough valuable parts to sell. To anyone, including current customers. The company failed to develop a strong portfolio of products they could “cross-sell.” Going back to Blackboard for a moment, however much people may grumble about them (rightly and/or wrongly), when the conversation shifts to the topic of their Ally accessibility product, eyes light up. It solves a real and important problem. People want it. Current Learn customers want it. Prospective Learn customers want it. Customers who wouldn’t touch Learn with a 10-foot pole want it. Therefore, it has financial value.

Instructure needs a portfolio of offerings that are compelling as Blackboard’s Ally in order to be financially viable in the long term. Because the company has thus far failed to develop such a portfolio, it developed an alternative growth strategy to sell an LMS into the corporate market instead. And not its existing LMS. No, a new one. Bridge.

This, frankly, was a dumb idea that was never going to work. Where the LMS is a mission-critical application that inspires passion—positive or negative—from users in the education markets, it is the polar opposite in the corporate markets. Whenever corporate budgets get slashed, Training and Development is the first departmental budget to get hit. And unlike the academic LMS market, there are somewhere in the neighborhood of 173,000 corporate LMS products. OK, I’m exaggerating. A little. But it’s a large enough list that whole companies have been built on writing up catalogs that describe the different LMSs in that market. Not even reviewing them, really. Mostly just listing them all in one place.

Ask yourself how detrimental it would be for you if Instructure sold off or killed off Bridge. If your answer is “not very,” then you don’t have much to worry about in terms of new owners selling off pieces of the company.

It’s important to understand that a profit motive and a motive to serve students and educators well are not inherently opposed to each other. They can be. They are almost always in tension with each other. But if the best way to make money is to serve students and educators well, then good results for education can be driven by a profit motive. Money will be more likely to be invested in the right things. I say “more likely” because the other challenge is knowing how to invest that money such that it will be likely to have the net effect of serving students and educators well. My father likes to say, “Never attribute to malice that which simple incompetence can explain.” More on this later in the post.

The PE acquisition would be to support a return to a focus on education

One way to read the PE acquisition is as an admission of failure of the Bridge strategy. That is certainly, explicitly, the line that Instructure’s activist investors have been pushing. If you’re worried about the impact of the short-sighted view of PE, then you should also be worried about the quarter-to-quarter pressures of the public markets. Instructure’s Bridge strategy failed to produce the financial results that were promised. One argument for PE acquisition in this situation is that the public markets would not be patient with Instructure as it attempts to recover from this misstep and build the product portfolio that it should have started working on five or more years ago. The right PE firm would give Instructure time to ditch Bridge and refocus on its core market. (That’s you.)

I don’t know much about Instructure’s PE acquirer, Thoma Bravo. Phil describes their investment style thusly:

Thoma Bravo is widely known for the “buy and build” acquisition strategy, where a platform company with solid customer base is purchased (often for high price), enabling subsequent acquisitions of smaller companies that have lower price multiples. This is not the same as buying company A and B and combining them; rather, this strategy is based on multiple acquisitions tied to the platform company.

That fits with what Instructure needs to do to become more sustainable in its core education markets. It needs more educational products to sell. So at first blush, Instructure being owned by Thoma Bravo could be better for customers than Instructure being publicly traded has been.

Interestingly, though, the current Instructure shareholders who oppose this acquisition also argue that Instructure should return its focus to its core market of education. They just don’t think this particular deal is the best way to do it and are suspicious that there are other, more venal reasons for this particular merger at this particular time. Here, for example, is what Instructure investor Rivulet argued in its SEC filing protesting the merger:

We invested in Instructure because we admire the company, and we see plenty of opportunity for continued growth and success. With Canvas, Instructure has developed the market-leading learning management software product for the higher education market. The Canvas story has provided an incredible lesson about the success that comes from a focus on innovation, putting students first, and offering a compelling alternative to a greedy incumbent that was run for the enrichment of management and shareholders. Unfortunately, with the announced sale of the company to Thoma Bravo, Instructure is providing a different sort of lesson to its customers and shareholders. Here, it appears that the company has run a rushed strategic review process that was designed to result in a sale to management’s chosen buyer at a low price…in order for management to save their jobs and enrich themselves in the years ahead. It is outrageous.

Rivulet SEC filing

In other words, investors both in favor of the merger and opposed to it agree that Instructure is a growth company that has stalled because it has lost focus on its core education markets. They believe the company can continue to serve its shareholders well by serving its education customers well. The debate among them is over whether this deal was really designed to serve those customers (and thus generate profit for shareholders), or whether it was corrupted by an agreement that exchanged preferential treatment of Thoma Bravo in the bidding process for particularly rich compensation for Instructure CEO Dan Goldsmith and his sister, who Goldsmith hired as a senior executive of the company.

I will not write about the details or the merits of those accusations in this post. That’s not e-Literate‘s beat anymore, and I have no insights to offer about them that would add any value. If you want to read more insightful and in-depth coverage of that fight than I could offer, then go read Phil Hill’s coverage.

Rather than speculating about motives or analyzing the finances, I am going to write about strategy and execution. Does Instructure’s current executive team have a credible strategy and a strong track record of execution that is consistent with its previous, incredibly successful strategy for growing and retaining its installed base? And does it have a credible, market-validated strategy for increasing the value that it provides to its customers such that customers will likely be willing to pay more for that value? If the answer to these questions is “yes,” then current shareholders can have increased confidence that they will receive a good price for their shares while the end users of Instructure’s products and services can have increased confidence that they can continue to count on the company to serve them well.

Unfortunately, it looks to me like the answers to these questions are “no.”

Why Instructure has succeeded—and failed—to date

We have long argued at e-Literate that Instructure came to market with three critical advantages that incumbent competitors were slow to see and which were difficult, time-consuming, and/or expensive for them to replicate. Those advantages are as follows:

1. Reliability (implemented as cloud-native infrastructure)
2. Ease of use (implemented as a rethink and redesign of existing LMS functionality)
3. Customer rapport (implemented through a variety of formal and informal ways and tied together by a strong company culture)

These three advantages can be collected under one unifying strategic insight of the original executive management team: They brought a consumer software sensibility to what had been treated by incumbent vendors as an enterprise software business.

Instructure understood who their customer really was in a changing market

For readers who are not software industry aficionados, one way to get at this difference is to ask yourself a few questions about software products being used at your institution that are generally treated as enterprise software products. How involved were you in selecting your registrar software? How broadly and intensively were different stakeholder groups involved in the selection process? Who made the final decision? Do you even know the answers to any of these questions? Do you know how the decision was made?

Generally speaking, registrar software is part of a larger package of software that includes modules for things like payroll and expense tracking. There is usually a very small group of people who both make the decisions on the procurement of such software and who also decide who else will have input on the process. This is typical with enterprise software selection—even for critical enterprise software that has many end-users who depend on it and use it daily. For enterprise software companies, their customers are those few decision-makers. They only have to care about reliability, ease of use, and end-user happiness to the degree that the decision-makers do. And often, those decision-makers have other priorities, like cost, or arcane but important support for regulatory requirements, or making their bosses happy in some way.

In the early years, LMSs were procured like enterprise software. CIOs generally selected the LMS vendor and decided on whose input they would take when making their selection. But by the time Instructure came on the scene, that was changing. Faculty were much more intensively involved with LMS selection, and it was not unheard of for a university to form a student advisory committee or conduct focus groups as part of the selection process. Instructure made and kept three promises to these stakeholder groups. First, they would build and run the LMS such that it would be extremely unlikely to go out for two weeks at the beginning of the semester or in the run-up to finals week. Second, it would be less painful and time-consuming to use than the alternatives which existed at the time. And finally, end-users would be treated as human beings, with good customer service and both visibility and input into the product roadmap.

Given the pedigree of the executive team at that time, these changes should not be surprising. CEO Josh Coates’ previous gig was running Mozy, a startup that was essentially the precursor to Dropbox. Think about what life was like trying to back up or share locally stored files across multiple devices before Dropbox. There was FTP. There were email attachments. And there was Sharepoint. But in order for a Dropbox product to replace those well-entrenched incumbents, it had to be rock-solid reliable, be dead simple to use, and feel completely approachable and safe.

Instructure has been spectacularly successful not because it has been a great education company but because it has been a great consumer software company. And the early management team was very clear about this distinction. Here’s what I wrote in a 2012 post:

The founders of Instructure aren’t educators, and they think that’s a good thing. The way they see it, educators tend to build software that solves their own problems from their own perspective, and the results are often idiosyncratic. As technologists coming in with no strong opinions about how teaching “should” be done, the Instructure leadership feel they have an advantage of humility and open-mindedness when considering solutions. As CEO Josh Coates put it, educators “are experts in their context. But their context is one or two classes in a specific university in a specific part of the country. Software developers don’t even pretend to know the domain.” So they have to go out and do the research. Instructure co-founder Brian Whitmer added, “And we know that we have to do it. We know that we have to validate it against a bunch of different people.” And they do. Instructure, as a company, is extremely attentive to the conversations among educators, to the point where they have Twitter feeds from various ed tech folks sucked right into the IRC channel that all their developers use for internal communications. They do their homework.

Educators tend to get hung up on the profit motive and, ironically, miss the disciplinarity of starting a company. Co-founder Devlin Daley talked about a “new style of development” that was “not true in the ’90’s,” and all three of them talked a lot about Agile development. As they see it, a lot of companies that implement Agile miss the fundamental aspect of the methodology that is about getting closer to customer needs.

e-Literate, “What Are Ed Tech Entrpreneurs Good For?“

Instructure lost its way after solving the obvious problems

In a lot of ways, Instructure got a big leg up by being late to the market. The product category was well defined through the experiences of the early incumbents. Does an LMS need a gradebook? Definitely. A test engine? For sure. A blogging system? Ehh…. Instructure’s product team could go to a lot of people who were already using LMSs and ask them, “What do you hate about this thing and what are you really trying to accomplish that is possible but painful in it?”

And, in fact, that is exactly what Instructure did. They figured out what problems people were trying to solve when they used an LMS and then, rather than building yet another, more feature-packed version of the same tool, they set out to build solutions to those problems. For example, they figured out that instructors who were entering grades wanted to do so as quickly as possible and focus their energy on adding educationally useful comments. So rather than cramming more features into the same spreadsheet-style gradebook that every other LMS had implemented, they built SpeedGrader, a novel user experience that helped instructors focus on the aspects of grading that they cared about while moving other functions out of the way. Why should instructors have a grade curve-setting widget taking up screen space when they’re trying to comment on a student’s paper?

But once Canvas had caught up functionally with the other LMSs, Instructure seemed to run out of ideas. Internal to the LMS, we haven’t seen many truly major customer-facing improvements in the past four years. In fairness, Instructure claims to have been doing a lot of under-the-hood work that will yield future benefits. But I’m still fairly unclear on what those future benefits will be.

Meanwhile, what new customer problems have they solved outside of the LMS? What new products have they given us?

Arc. They gave us Arc. A lecture capture tool. Another lecture capture tool. Is it good? Yeah, it’s pretty good. Did it set the world on fire? No, it did not.

I don’t want to overplay the we-don’t-need-another-one-of-those card, because Canvas itself could have been dismissed on those grounds when it first came out. In fact, it was dismissed on those grounds. I, personally, dismissed it on those grounds.

Nor is it the only example of such a product. Not by a long shot. Consider Zoom. It dominates the webconference market, both in education and elsewhere. How did that happen? Cisco already had a very mature and very widely adopted product in Webex. Google had Hangouts. Adobe had Connect. Blackboard had the education-specific Collaborate solution. There are multiple open source options, like Big Blue Button. There are other options like GoToMeeting and Bluejeans. There was no good reason to believe that a startup with yet another webconferencing solution would survive, nevermind thrive.

And yet, it did.

Why? This is the critical question for Instructure to both maintain its current strength and build new ones. Maybe Instructure just got lucky by coming in when they did. Maybe they weren’t nearly as talented as they seemed to be. Humans tend to over-credit skill and under-credit timing and luck.

Even so, I don’t believe that luck and good timing were the primary reasons for Instructure’s success. Permit me to repeat a snippet from the quote above of my 2012 post:

Co-founder Devlin Daley talked about a “new style of development” that was “not true in the ’90’s,” and all three of them talked a lot about Agile development. As they see it, a lot of companies that implement Agile miss the fundamental aspect of the methodology that is about getting closer to customer needs.

I believe Instructure had an approach to building their business that worked. They lost some mojo because, among other reasons, the pressure for them to grow pushed them into flattening what had been a philosophy and corporate culture into something closer to a paint-by-numbers formula. This is one reason why sales have slowed. Instructure is no longer performing as a truly great company. And if they don’t find their mojo again before they enter into an acquisition, then things are far more likely to get worse—possibly catastrophically worse—than they are to get better, because building a culture of product development excellence is not one of the things that PE knows how to do well. Whatever patterns are there when the company is purchased are likely to stay. Whatever direction the CEO has set will stay. And without the bulwark of leadership protecting a strong culture of excellence, the tendency to paint by numbers is likely to get worse rather than better. PE firms may or may not understand education, but they all understand numbers and find a high level of comfort in them.

Product-led Growth

Back when Instructure was being started, the software product development buzzwords were “Agile” and “Lean.” Today, a new buzz-phrase that also fits with that quote above is “product-led growth.”

“Product-driven company” can be defined narrowly or broadly. The narrower version involves some fremium model where people adopt a free version of the product through a self-service process, use it a lot and, once again through self-service with no human salesperson involved, upgrade to a paid version. Zoom works like this. So does Dropbox. I had a free Dropbox account for a long time. To borrow a phrase from Apple, it “just worked.” I used it to the point where I needed more space. So I paid for an upgrade. At peak Dropbox for me, I had a personal paid account as well as a larger one for my company.

Spreadsheet jockeys like product-driven companies because they have low sales and marketing costs. People adopt the product because it’s easy to do so. It spreads because people tell their friends and colleagues about how good it is. It makes money because people love and use the product so much that they will pay for an upgrade. The customers do much of the sales and marketing work and pay for the privilege.

But as with Agile, people who implement a product-driven company design miss the fundamental aspect of the methodology that is about getting closer to customer needs. And again, Josh Coates’ last gig before becoming Instructure’s (previous) CEO was building the precursor to Dropbox. There wasn’t a big self-service component to Canvas sales, but the core philosophy of creating a customer experience that carries a lot of the sales and marketing load by inspiring fanatical loyalty was there.

At least, it was there in the beginning, and for quite a while. It has been less evident in recent years. When a company is under pressure to grow, either financially or just to scale up to meet the needs of new customers that are walking in the door, it is easy to slip from a focus on finding new problems to solve toward one of finding new things to sell. In EdTech, there aren’t many obvious things to sell, particularly “at scale.” Painting by numbers under these circumstances is a weak strategy that fails often.

But education has many, many problems to solve. If you have a strong rapport with your customers, a really solid team of education specialists who understand the complex nuances of educational problems, and a leadership team that is absolutely committed to growing by understanding their customers’ needs better than anyone else does, then you can find ways to make money by solving new problems. Product-driven companies are problem-driven companies.

A good example of solving new problems is Blackboard’s decision to acquire Ally. Was there a healthy, established product category for an educational content accessibility checking platform that would be sold to colleges and universities? Absolutely not. Could anyone calculate the total addressable market (TAM) for such a product without wildly guessing? Not in any way that I can think of. Were Blackboard’s customers begging them to build and sell such a product? I very much doubt it, since there really wasn’t any such thing on the market.

But Blackboard saw that the little startup that had found this problem to solve. I’m sure they talked to Ally’s customers—and their own—about it to make sure that it was a real and large problem, solved by the product in a useful way. And they made a bet. It’s turning out to be a good one.

In retrospect, you could argue that the product is adjacent to Blackboard’s LMS business. But almost everything in EdTech is adjacent to the LMS because almost everything has to connect with it or work with it. If you had grabbed a random EdTech mid-level manager off the street and asked them for their top suggestions for hot product categories that are adjacent to the LMS, I doubt that learning content accessibility would have shown up on anybody’s list. This is the kind of thinking that made Instructure, and it’s the kind of thing it needs to be doing again if it wants to find its next leg of growth while holding off its newly resurgent competitors.

So who did Instructure’s board hire to lead the charge at this critical moment to reinvigorate that deep culture of devotion to solving the problems of the end-users? A guy whose last job was running a company that sold enterprise software to life sciences companies.

Now, Dan Goldsmith is not his résumé. He’s a multidimensional person who is capable of learning. (And to be fair, Josh Coates was not an obvious candidate for the job either.) But Dan has been in his current job for less than two years. In order for him to establish that he is up to this challenge in this industry within that length of time, he would have had to perform extraordinarily well. This is vital right now because the very same shareholder ballot they are voting on to accept Thoma Bravo’s offer would also change Dan’s compensation package to one that would attempt to lock him in with a very generous amount of stock that vests over time. If Instructure’s board and shareholders have decided that now is the time to go private, with this CEO, then they need to be very confident that he will deliver. If they are not, then the customers are under increased risk, the new owner is under increased risk, and the current shareholders are less likely to get attractive competing bids for their shares.

Has Dan Goldsmith met this high bar in the brief 20-month period that he has been on the job?

In my opinion?

No.

A case study in the opposite of what Instructure needs to do

Instructure hasn’t made a lot of significant moves in the education space since Dan took the helm. Off-hand, I can think of two. The first was the acquisition of Portfolium. In and of itself, that’s not an obviously genius move. The ePortfolio product category has been around almost as long as the LMS product category and has been a hot seller since…uh…never.

It could turn out to be an interesting move. Instructure has to bet that (a) potential trends like comprehensive learner records or stackable credentials will turn into something more substantial and that (b) an ePortfolio in general and this ePortfolio, in particular, provides a good infrastructure from which to start supporting those trends. I’m troubled by the fact that I’ve not heard an articulation from the company about their vision for the product. Dan made some comments about the potential for cross-selling. But those comments characterize Portfolium as a thing to sell, not as a solution to a customer problem. I simply don’t know what they’re thinking with Portfolium.

The other move was Dan’s announcement about their DIG learning analytics initiative in the spring of 2019, which Phil covered in his own inimitable way:

The second initiative announced on the earnings call was DIG, a strategic move with data and analytics.

“I am also pleased to share with you an early insight into our second growth initiative focused on analytics, data science and artificial intelligence. The code name for this initiative is DIG. And this technology platform combined with the most comprehensive SaaS database on the educational experience uniquely positions us to deliver meaningful value to our customers. And from a growth perspective, DIG has the potential to double our TAM in education.”

Instructure started ramping up their data and analytics efforts (again) about a year ago, although the focus was described at the time as being about internal analytics – that is, making Canvas a better and more valuable LMS product. From what I have heard the product validation for DIG are consistent with this message – dashboards, surfacing useful data within a workflow, etc. But that was not how DIG is being sold during the conference call [emphasis added].

“We’ve been working on the scaffolding for [DIG] for well over a year now. I mentioned in our remarks that we already have product validation towards out there in the market. We have instructors and students consuming output from some of the initial experiments with DIG. And we anticipate later this year obviously to make more announcements around specific products and offerings and how we bring them into the market. DIG ultimately is a platform first and foremost based upon machine learning and artificial intelligence. I believe that any multi tenant SaaS company born in the cloud has the opportunity once they hit a certain market share. And in fact, it may even be incumbent upon those organizations to partner with the industry and evolve that industry with new insights and predictive modeling using AI and ML. That’s what DIG is at its heart.“

This is brand new behavior for Instructure as a company. Previously the company was reticent to talk much about non-released products, but now they are talking not just about a new initiative, they are touting buzzwordy machine learning and artificial intelligence and predictive modeling well before any of those capabilities exist or are in customer hands. Goldsmith further clarified the DIG plans during the investor conference discussion [starting at 9:00, emphasis added].

“We already have analytical capabilities in our Canvas platform. I want to be really clear and delineate the difference between an analytics and reporting capability, and a machine learning and AI platform. [snip]

“We have the most comprehensive database on the educational experience in the globe. So given that information that we have, no one else has those data assets at their fingertips to be able to develop those algorithms and predictive models.”

Goldsmith then described an example of predicting a student’s expected performance in a class and how that prediction reliability goes up over time. Then we get the vision.

“What’s even more interesting and compelling is that we can take that information, correlate it across all sorts of universities, curricula, etc, and we can start making recommendations and suggestions to the student or instructor in how they can be more successful. Watch this video, read this passage, do problems 17-34 in this textbook, spend an extra two hours on this or that. When we drive student success, we impact things like retention, we impact the productivity of the teachers, and it’s a huge opportunity. That’s just one small example.

“Our DIG initiative, it is first and foremost a platform for ML and AI, and we will deliver and monetize it by offering different functional domains of predictive algorithms and insights. Maybe things like student success, retention, coaching and advising, career pathing, as well as a number of the other metrics that will help improve the value of an institution or connectivity across institutions. [snip]

“We’ve gone through enough cycles thus far to have demonstrable results around improving outcomes with students and improving student success. [snip] I hope to have something at least in beta by the end of this year.”

Wow. Robot tutor in the sky – meet the new kid on the block.

For those readers who don’t follow the sector broadly, Phil’s reference to a “robot tutor in the sky” is from an infamous quote by the CEO of now-defunct Knewton that they liked to think of their product as “a robot tutor in the sky that can semi-read your mind.” In other words, Phil is accusing Goldsmith of using tone-deaf language that would be explosive to educators. And sure enough, it pissed a lot of people off. They didn’t like the vague hype, and they didn’t like the seemingly out-of-the-blue implication that students’ data from their in-course activities would be used without their permission for Instructure’s product development purposes. As I mulled whether and how to write this post, I received the following reply on Twitter to my retweeting of one of Phil’s posts on the topic:

Goldsmith: "operating in the public spotlight wasn’t fueling innovation…"
Instructure has not been very transparent about "Dig" and data mining for AI/ML projects so far and I guess that's going to get way worse; see Goldsmith's letter as quoted here in Phil's helpful analysis. https://t.co/BLgNVWSNGC

— Laura Gibbs (@OnlineCrsLady) January 2, 2020

OK, OK, so somebody on the internets wrote a mean thing about Instructure’s DIG strategy. And is still upset nine months after the announcement. Maybe that evidence doesn’t persuade you (particularly if you don’t know who Laura Gibbs is).

How about this?

I've drafted a letter for higher ups regarding concerns with Instructure & the unknowns with student data. Would love some feedback if you have a minute, esp those who've spoken up and/or directly affected by this! Doc is open for comments. https://t.co/tG2CZ74atd

— Cristina Colquhoun (@call_hoon) December 14, 2019

That’s a Twitter thread from Canvas end-users who are crowdsourcing a letter encouraging their institutions to lobby Instructure shareholders about Instructure’s lack of legally binding data privacy policies. Now, the authors had voiced a range of concerns in the last draft (plus comments) that I looked at before writing this post. Some of them may have been poorly phrased, overblown, or factually questionable. But others were unquestionably legitimate, sophisticated, and serious.

Update: The authors of the letter have posted the most recent draft here, and a signable petition here.

I see three big take-aways from this letter which are most important for our current purposes. First, this is the first time in my long history of covering EdTech that I have ever seen educators undertake such an effort to influence shareholders. The last company I would have expected to inspire this level of concern-driven activism is Instructure. And the existence of it strikes me as one of the most singular and remarkable milestones of Goldsmith’s tenure to date. Trust me: This should be taken seriously.

Second, one reason this foment continues to build nine months after Dan’s original comments about DIG is that, at least to my knowledge, Dan himself has yet to address these customer concerns directly, publicly, and credibly. Even in this Twitter thread, he leaves the job to an anonymous PR person to desperately try to tamp down the anger. At this point, given what’s at stake, he should be making a direct personal statement as the CEO of the company. If not on Twitter, then on the company blog. Or somewhere. Anywhere.

Finally, the authors of the letter make clear they understand that there are no obvious, well-trodden solutions to the student data privacy conundrum. They request some specific steps at the end of their letter, but they don’t pretend that these steps solve the complex, pervasive, and important challenges that universities who work with EdTech products face. They are struggling to balance the need to protect student privacy against the affirmative ethical obligation to learn how to better help their students succeed.

THIS IS A PROBLEM THAT CUSTOMERS NEED HELP SOLVING.

Instead of either assuming a defensive crouch about DIG or ignoring complaints altogether, Instructure could be listening to customer concerns with an ear toward making money by serving them better. This is a hard and important problem space that will not be solved without active participation by vendors. There have been academic convenings on the student data privacy challenge by groups like Asilomar. They do critical work. But they are not enough. There are initiatives by industry groups like IMS. They too are necessary but not sufficient. This is an important and growing problem, which makes it an important and growing commercial opportunity for a vendor that can actively partner with academia in finding a solution to it.

Dan Goldsmith could have easily turned his faux pas into a moment of leadership. There is even precedent from Instructure for this under strikingly similar circumstances. I encourage you to (re)read Phil’s 2015 post on how Josh Coates handled customer complaints about the company charging universities for access to their own students’ data. The short version is that once Josh became aware of customers’ unhappiness, he took it as an opportunity to engage with them, changed the policy, published a mea culpa post on the company blog, and encouraged Phil to independently verify with the customers that their concerns had been addressed.

Dan’s comments were significantly less serious than the implemented commercial practice that Josh had to correct. He could have easily addressed them. And there is a real problem space here that cries out for product development. EdTech needs an architecture of privacy. Instructure’s customers are being very direct about their needs and concerns. The company even has access to multiple academic experts who could help them think through the nuances of the problem. I know this for a fact because I introduced Instructure employees to those experts last spring when I urged them to turn the DIG situation around by taking leadership.

That was over half a year ago. And now we have customers banding together to lobby shareholders about their continuing dissatisfaction. Because apparently, they have come to the conclusion that the company leadership is not listening to them.

No. I would not bet the company on the 20-month record of Dan Goldsmith’s leadership and articulation of strategic direction. Maybe things will be different six or twelve months from now. But at this point, he has neither embodied nor provided a sound investment thesis. There is no strong reason to believe that Instructure, sold to private equity under this leadership at this moment, would be able either to continue to grow market share of its core product or to develop new solutions to real customer problems that would lead to greater profitability.

Just how bad it could get

I want to wrap up this post with a couple of hypothetical but plausible examples of just how bad things could get when an LMS company with no products to cross-sell and no customer-focused, problem-solving leader sit down in a room with PE owners who have big ambitions, a lot of money, and limited experience in the space. A player like Thoma Bravo is going to make acquisitions. If they make a number of small bets that don’t pay off, then there are likely to have to cut costs and raise prices to make up for their losses.

But what I really worry about are the kind of colossal mistakes that can happen when a CEO who doesn’t know education and investors who don’t know education lock themselves in a room together and try to come up with something really big. Phil noted in a recent post about Thoma Bravo,

Another point in the Forbes profile is that Thoma Bravo does not acquire a company and then figure out what other companies to tuck-in or combine [emphasis added].

“Like a good tennis player who’s worked relentlessly on his ground strokes, Bravo has made private equity investing look simple. There are no complicated tricks. He figured out nearly two decades ago that software and private equity were an incredible combination. Since then, Bravo has never invested elsewhere, instead honing his strategy and technique deal after deal. He hunts for companies with novel software products, like Veracode, a Burlington, Massachusetts-based maker of security features for coders, or Pleasanton, California-based Ellie Mae, the default system among online mortgage lenders, which the firm picked up for $3.7 billion in April. His investments typically have at least $150 million in sales from repeat customers and are in markets that are too specialized to draw the interest of giants like Microsoft and Google. Bravo looks to triple their size with better operations, and by the time he strikes, he’s already mapped out an acquisition or turnaround strategy.“

Instructure fits into this playbook, and I suspect that we won’t have to wait long to see the next move. It could be tuck-ins in the vein of Portfolium, Practice, or MasteryConnect, or it could be something bigger and market-changing. It is worth noting that Bravo raised $12.6 billion in January of this year – the stakes are getting higher.

What would fit into the category of “bigger and market-changing”? I can think of only two plausible acquisitions, particularly since Thoma Bravo seems to be disinclined to invest in people-heavy businesses like OPMs or textbook publishers. Either one of these ideas would be so incredibly, mind-bogglingly stupid that I’m hoping this part of the post is just good for laughs. (Or maybe a Halloween-style scare. Boo!)

The first possible merger target is Ellucian which, as Phil has pointed out, has put itself up for sale. On the spreadsheets, this could look like a great prospect for Instructure. They sell to the same client base, have a big installed base, a portfolio of products to cross-sell, and have a massive lock-in.

Anyone who actually knows EdTech and is thinking from a product-led perspective would see this as the obviously terrible idea that it is.

In general, the SIS market is a smoking ruin that will take at least another half a decade to move itself into the cloud. The technological heritage and history of the product category conspire to make Ellucian—and their competitors—about the furthest thing from a product-led company possible. Think about how hard it was for Instructure’s competitors to make the transition they needed to make in order to catch up. Now multiply the challenge by a kajillion. Combining Ellucian with Instructure would be a disaster.

The only worse idea I can think of that might somehow seem plausible to the wrong people having the wrong conversation at the wrong time would be for Instructure to buy—are you ready for it?—Blackboard. Again, from a paint-by-numbers perspective, it makes perfect sense. Instructure could buy Blackboard’s remaining customer base very cheaply, along with its more successful portfolio of cross-selling products. (Their Ally accessibility platform, for example, is a well-deserved smash hit.) And since Blackboard is close to having zero financial value, with their debts being close to the intrinsic value of their assets, the sellers could unload an investment that is no longer performing for them while the buyers could get assets they need practically for free.

And in the process, Instructure would acquire the worst brand in the product category—which is an improvement from when they were the worst brand in the history of the sector—by applying the most infamous and most loathed business strategies of their infamous former Blackboard CEO Michael Chasen—namely, buying up one of the few competitors in the market. Instructure could literally become the new Blackboard. In doing so, they would utterly destroy one of the greatest brands in the history of the sector instantaneously.

These guesses must be wrong. (Dear Lord, please let them be wrong.) Think of them as hypothetical examples of “really, Michael, how bad could it get?” than as actual predictions. I have to believe that the board, the executive team, and Thoma Bravo would not be that dumb. But the point is that of those three entities—the board, the financial owner, and the CEO—the only one we would reasonably expect to know how bad such decisions would be in practice, the only one that all parties must be able to count on to steer the ship away from the many rocks and icebergs in EdTech, is the CEO.

Has Instructure provided customers with a detailed and credible enough strategic roadmap to inspire confidence that they have a more compelling alternative for growth? No, they have not. Has Dan Goldsmith thus far proven, lacking such a roadmap, that his reputation for performance alone is worth betting the company on? No, he has not. No smart PE company would make an attractive counter-offer under these circumstances. There is no sound investment thesis until Instructure is able to regain its footing as a product-led company.

Instructure has been one of the best run, most consequential companies in the history of EdTech, and it could be again. It has not yet experienced the talent flight that would make a turn-around much, much harder. But until it has both a credible strategy and a leader who inspires customers to trust the company with their most pressing and complex problems—like fulfilling their ethical obligation to protect their students’ privacy in an era of digital learning—it will be just another EdTech company that is toiling away in an unprofitable product category and promising investors that it will somehow spin straw into gold.