Investor Sentiment: 2025 Ed Tech Market Perspectives

Garrett: At a high level, the evaluation drivers and performance indicators in the ed tech ecosystem are really no different than in the broader software and tech ecosystem: a continued focus on profitable growth as measured by the rule of 40, a strong focus on retention (particularly gross retention) and the market opportunity for companies with large TAM runways.

For ed tech specifically, there can sometimes be a question as to how much room there is to run, especially in more narrowly focused or more niche subsegments of the ecosystem. Making TAM focus a particular importance for ed tech is really what I would highlight.

While it ultimately depends on your business model and company’s positioning, there has been an increased focus on the funding model of the customer over the past 2 to 3 years that’s unique to the current ed tech environment.

Is the customer buying your products using public funding, or some other funding source that is exposed to budget cuts or removal? And what is the actual health of the institution or customer you’re selling to?

Are they facing strong increasing enrollment trends or are they facing their own headwinds and decline?

Given that potentially uncertain funding backdrop, from an investment standpoint, there is an increased focus in software and technology that delivers high and hard dollar ROI to its customers, with real quantifiable impact on their businesses.

Thus, the ed tech vendors that are selling with a high ROI story are commanding better valuations from the investor community.

Garrett: I’m a big believer in the long-term potential in this category. There will always be ebbs and flows in certain segments and subsegments, but the macro trends are really quite strong here when you think about:

• What technology is able to deliver in terms of better efficacy and adaptive personalized learning pathways that meet the student where they are and deliver a better learning experience
• The number of learners who are in some sort of remote or virtual or distance environment that require digital ed tech infrastructure to deliver education
• The need for schools, institutions and corporations to better rationalize their operations and improve efficiency
• The need for continued training and reskilling and certification among the workforce

These elements all add up to a lot of macro tailwinds that will continue to be at the back of ed tech companies for the long term here.

Jeff: As Garrett said, I think that we all focus on that rule of 40, yet the composition of that rule of 40 has changed a little bit over the past several years. A couple of years ago, valuation was far more about revenue growth. In a zero-interest rate environment, you could spend a dollar to chase a dime, and that was looked upon favorably.

But nowadays, there is a little bit more emphasis on EBITDA. And while revenue growth is still important, previously overlooked elements like margins and the manner in which you generate that revenue, as well as the cost of acquiring a customer, are starting to come to the forefront.

With interest rates to where they are at now, people are starting (especially investors) to really want to understand how some of those revenue dollars are trickling down to the bottom line, more so than they did a few years back.

Garrett: In general, we see a pretty wide range of revenue and monetization models in the ed tech ecosystem. There is a high prevalence of volume-driven revenue models with metered-based billing that look at things like number of active students or active learners and the number of courses or assessments delivered, all of which can be a good thing and a bad thing in a highly-increasing volume environment.

They can be great because they are easy to embed with the same store customer growth with little to no effort on your part, but on the flip side, what goes up can also go down again as well.

What we have seen as the most attractive model, if you can find a way to construct it and pull it off, is a hybrid where you can protect your downside with a fixed subscription and and maybe a fixed allocation of volume, but also uncap the upside by having overage or usage-based charges over that fixed minimum.

And for companies that can really nail that balance, it can be an extremely attractive combination of both effortless growth from increased consumption by clients, but without the negative downside risk that typically comes with a volume-based model.

From a valuation perspective, a key thing to consider, or to focus on when calibrating your pricing model, is that you want to try and strike it so customers are consuming at or near their contractual minimums.

We have run into big challenges before with companies that have those minimums, where customers are actually consuming way below, which is a huge potential churn and attrition risk. Finding that balance can be tricky, but if you can pull it off, hybrid pricing is a really attractive option.

Jeff: My perspective is going to be more broad-based, beyond just ed tech. If you just look at all the verticals, I think everyone is going through a very similar pattern where, again, years ago, SaaS or that fixed license-based or subscription-based model made a lot of sense, and it was worth a lot of money.

In those days, you didn’t really need to think too much about EBITDA, and the cost model supporting such a subscription fee wasn’t understood to a large, detailed degree. 

Now people understand that there is a lot of price sensitivity across various segments, regions and customer bases. As a result, you see an increase in regionalized pricing models and more bundled pricing options.

And as folks start to package things together, and as their operations grow more mature, they can entitle features in more of a sophisticated way to bundle and gate things, and as a result, their subscription models start to look a little more sophisticated.

And what we’re seeing now is a movement towards outcome-based pricing. You see the Salesforce and the Google’s of the world that are starting to come out with price-per-resolution in the support industry, which is going to start moving consumers towards looking at pricing from that vantage point, and obviously, there are pros and cons.

The more sophisticated and variable you get, the less predictable it could be from the customer side. So if CFOs are your buyers, they can struggle a little bit with [the variable aspect] of an outcome-based pricing model because they want predictable revenue.

On the flip side, consumers like a dynamic pricing model because they are more willing to pay for an outcome they value rather than just paying fixed pricing. And as Garrett mentioned previously, striking that balance is the holy grail – that fixed concept of covering costs and baking in margin, along with that variable component that can free up upside.

At the same time, however, you have to understand the margins that you’re achieving with each of these items given that’s what investors are really scrutinizing now.

So that’s where I see everyone is going, and as a result, I’m seeing a lot of companies doing price testing. Where customers might have changed their pricing once every couple of years, I’m now seeing folks introduce changes to either their packaging or pricing multiple times a year.

Garrett: It’s a topic that is clearly very top of mind, and one that has a lot of different nuances and facets to it. To break it down, let’s think about the landscape in two broad buckets. In your first bucket, you have your incumbents with existing big business models delivering something of value to the customer base.

In your second bucket, you have your new start-up, AI native companies that are building disruptive products to come after those incumbents.

From an evaluation perspective, I think for the AI native players, those same fundamental metrics that we were talking about earlier – revenue quality, revenue growth, unit economics, rule of 40, gross net revenue retention – all still apply.

For the incumbents, the answer is actually a lot more nuanced. I think those players in the business today really need to take a hard look to assess both the revenue potential and the competitive threats to their existing business.

We have seen a lot of opportunity with larger incumbents developing new AI products to drive continued and increased revenue growth. We have also seen a lot of potential for them to adopt AI internally to drive down costs, improve efficiency and drive up profitability.

At the same time, there is that disruptive threat on the horizon from those new AI-native entrants that I just alluded to.

The big questions for an incumbent now are, “how does the emergence of AI in my business impact the status quo? How will my customers change their behavior? What competitors will likely emerge or are emerging? What opportunity can therefore be created in my own business by AI?”

It’s not a one-size-fits-all answer, but I think a pretty holistic review of the landscape is probably required for everyone. Overall, most of our clients and the companies we are talking to are embracing AI as an opportunity rather than worrying about it or seeing it as a disruptive threat.

Jeff: We have been engaged with a number of our customers on this subject. Not to say that this is the one-size-fits-all answer, but here is the common approach we have seen:

We had a customer who developed a co-pilot feature and gave it to their customers for free so they could start to see adoption and usage patterns. We partnered with them and came up with a plan to, based on the usage patterns, start charging customers, and see what that would do to their adoption rate and usage patterns.

They started by first building a fixed price per user type of license, came back to us, and we looked at their usage patterns together. They were able to provide answers to questions around how their customers were consuming things and using this co-pilot feature.

Then we determined that it was time to start setting some thresholds to test their pricing. They looked at one segment of their customers and charged them one rate, while they took another segment of customers to test an outcome-based pricing model.

Once they started getting data from those price tests, we started comparing those different pricing strategies to determine what got better free-to-pay trial conversions. And from that standpoint, once they understood the cost and the values, they started thinking through what Garrett mentioned earlier – a hybrid pricing model. 

So you can see how this process first starts with, “how do I look at the data? How do I see usage patterns? How do I understand from those usage patterns, the value and cost vectors?” And most fundamentally – “How do I experiment and conduct price testing?” 

Once you implement a structure that sets you up to answer those questions, you’re in a position to report on some of these key pieces. Those have been the common steps that we’ve helped Chargebee customers go through. You can see that it’s not a “let’s just go out there with a huge launch!”

It’s a “let’s dip our toe in the water, understand what we’re seeing and start experimenting with monetization strategies thereafter.”

Now, to talk about outcome-based pricing. Supporting outcomes is easier in some industries than others. Think of resolution tickets. If a customer issue can be resolved, that’s an outcome. In the ed tech world, or for on the job training, it could be a completion of a certification, or something of that nature.

You’re trying to marry the usage to the outcome with the inherent cost of using/building on top of an AI tool. When you’re evaluating a usage based pricing model, you want to make sure you’re considering the cost of the AI component that’s powering that. The outcome you’re charging for has to consider your cost and the AI tool you incorporated into your solution.

Garrett: We recently worked on a transaction with a company that was, to use my prior categories, a successful incumbent, and they had a big initiative. They were spending a lot of money, time, energy and resources to develop new AI products, and a big part of the future growth story for that business was the ability to sell those new AI products to their existing customer base.

As we thought about telling that story, we tried to think through what makes this company’s ability to win differently than others that may come after the same opportunity.

Their biggest asset: they had a large existing customer base, and the nature of that large customer base gave them access to lots of real-world data to better train and hone the AI models that they were working on.

That existing customer base gave them the audience to embed what they called “moments of wonder” into their existing product, where they could expose some of that AI functionality to their existing users and use that to catalyze a conversation around more of a fulsome AI product upsell by giving the users a taste of what the AI could do over and above the core product.

That story was all generally very well received by investors, and the conclusion from many who evaluated that business was that there really was a right to win in and a compelling AI opportunity to drive future revenue growth and future value growth for the business.

What was crucial to their success was having well-thought-out answers to the question like  “how much of your offering is proprietary and what gives you the right to win?” and “what are the defensible notes around your AI strategy?”

There are a lot of investors who are wary of AI for the sake of AI type stories. So as you think about where to focus your AI resources and efforts, the best place to do it is where you have those real defensible moats and that right to win.

I’ll demonstrate something else to be mindful of as you are going through this AI product development journey with another example. There was a strategic buyer that was a large public company that had more stringent compliance requirements as it relates to the usage of customer data, particularly in the context of AI, that was looking to acquire one of our clients.

And that strategic buyer voiced some concern around how our client had been using existing customer data to build their AI products.

While ultimately it wasn’t a deal-breaking issue, the lesson learned here is that to get credit for a new AI product, you need to be very mindful of those underlying compliance and data usage implications if you’re hoping to attract interest from larger strategic type investors or acquirers who may have a higher bar in those regards. I’d encourage you to think about building your product with that higher bar in mind.

Jeff: There’s no real right answer for some of this stuff. You see that people like Salesforce and others that are coming up with new monetization and new AI, and the pricing is all over the place for many of these folks.

So no one’s really figured out exactly the right Goldilocks model. But based on my experience in the industry and with Chargebee customers, I have 10 topics of consideration that I would recommend if I were to launch something new:

The first topic of consideration is cross-functional sponsorship. Please don’t incubate this idea – this idea needs to be looked at from multiple perspectives. The CFO’s office needs to think about it from a monetization side.

The go to market folks need to think about it from the sales side and the product folks need to think about it from the data side. All of that needs to be looked upon in concert with each other. I have seen some companies fail by trying to get 12 people together to figure out an idea without really thinking through its cross-functional capabilities.

The second topic is what Garrett previously mentioned – market needs. It’s really understanding that right to play. What is it actually accomplishing for me? Is this a play to generate more revenue? Is it a play to retain customers? Getting a handle on what that market needs is so important.

The third topic is setting clear goals. Most customers I’ve talked to have set out short-term goals – first, I’m going do something internally, then I’m going introduce a co-pilot and look at adoption, and then in the long term, I’m going look at agentic AI and figure out how I can use agents within the process to to handle workflows.

So making sure there’s a clear understanding of goals, both adoption and financial, is also crucial.

The fourth topic is data and documentation. Documentation is extremely important when you’re doing stuff like co-pilots. They have to sit on top of proper documentation in order to understand your data and things like your usage patterns. You need to have the right level and quantity of data that allows you to get proper insights into value and costs.

The fifth topic is talent, and what I mean by this is – do you actually have the right people internally? Garrett mentioned in one of his examples of a customer using their own customer data to model behaviors, and things of that nature. That implies that you have the talent in-house to run analysis like that.

The sixth topic is understanding what talent you have in-house and what you could be doing internally versus leveraging external. There are lots of vertical models out there who have already done some of this stuff, and maybe you’re just going to leverage those, maybe you’re going to tweak those, for your particular niche.

But understanding what’s available externally and what you are capable of building internally is an important decision to be made. 

The seventh topic is the ability to figure out what drives your costs. For some companies, there may be a lot of cost considerations that are variable and don’t match up with your revenue producing side, which can get you underwater very quickly. So understanding the cost vectors is really important.

The eighth topic is considering your pricing and packaging structure. We’ve talked about different pricing models, leveraging both fixed, consumption-based and outcome-based, but all of those are going to drive some productivity, and you need to be able to understand the margins associated with it.

Once you understand those margins, then you need to understand how you’re going to package and link it to the current plans that you have.

The ninth topic is experimentation. Most companies are looking to experiment with this stuff, and not just experiment with the offering, but to experiment with the pricing and outcomes to determine how they can best maximize their profitability.

Lastly, my tenth recommended topic is the thing that often gets overlooked – operational execution. Once you have a great idea, how are you linking that and connecting all the dots with your entitlements?

How is that idea manifesting in your billing and collections? How are you accounting for all this stuff? All that stuff on the back end needs to be thought through for you to understand what type of margins and metrics you’re achieving through these programs.

Garrett: I think Jeff nailed it. I’ll just summarize his points with two points. The first – a strong answer to the question of what gives you the right to win is a key one to keep in focus as you think about where to prioritize your AI strategy, and second, from a long term evaluation perspective, keep in mind that eye towards the strong fundamentals we opened the conversation with.

It’s OK to make investments to get new strategies off the ground, but you need to have a profitable, growing, strong unit economic business model on the other side of it to make it all worthwhile.

If you enjoyed this conversation, or are currently exploring AI monetization strategies, the Chargebee team is here to help. Feel free to reach out to us here or subscribe to Raymond James’s quarterly newsletter.

Garrett DeNinno

Managing Director, Raymond James Technology and Services Investment Banking

Garrett DeNinno is a Managing Director in the Raymond James Technology & Services Investment Banking practice with a focus on the education technology and vertical software sectors. With nearly 20 years of experience, Mr. DeNinno offers deep expertise in the technology and services industry and has completed approximately 100 transactions. He rejoined Raymond James upon receiving his MBA from the Tuck School of Business at Dartmouth. Mr. DeNinno previously worked for Raymond James as an analyst and associate in the Technology & Services Investment Banking group after receiving his BA from the University of Virginia.

Jeff Sant

Chief Customer Officer, Chargebee

Jeff Sant, CCO at Chargebee, brings 27 years of experience in software entrepreneurship, serving in leadership roles across all aspects of the GTM motion. After 9 years in management consulting, he founded his first company and later co-founded Primatics Financial, a SaaS company focused on loan accounting, which he then sold to a large public company. From there, he was a founding member of RevLock, now Chargebee RevRec, ensuring its successful integration into Chargebee’s GTM process. Jeff resides in Northern Virginia with family, loves golf, tennis, the Steelers, and anything Italian.