How ClickUp blends PLG and sales-assist

COO Gaurav Agarwal on cutting CAC payback in half, the modern PLG org structure, and embracing incrementality over attribution

👋 Hi, it’s Kyle and I’m back with a new Growth Unhinged, my newsletter that explores the unexpected behind the fastest-growing startups. My new Icons series uncovers how top brands like Calendly and Miro build modern revenue machines. Up today: ClickUp.

---

You might recognize ClickUp for its Super Bowl ad, viral music album (over a million streams on Spotify), or path to crossing $100 million ARR in under five years. The business also happens to serve 10 million users (including 80% of the Fortune 500) and is #63 on the Forbes Cloud 100, a list of the best-performing private software companies.

I’ve long admired the all-in-one productivity platform as one of the rare software companies that has nailed a modern go-to-market (GTM) strategy combining both product-led and sales-assist. (I previously featured their innovative approach to pairing outbound and PLG.)

Gaurav Agarwal joined ClickUp as Chief Growth Officer in February 2022. He was recently promoted to COO as a result of a fantastic track record. Under Gaurav’s leadership ClickUp has:

• Honed its PLGTM distribution model, combining the best of Growth, Product, Sales, CX, and Data

• Cut CAC payback in half

• Continued to beat its annual plan and deliver top decile YoY growth

Product-led go-to-market (PLGTM) is the future of SaaS (and AI) distribution and go-to-market strategy. The goal is to combine the best of digital and human touch, and in doing so deliver a seamless customer experience that efficiently unlocks revenue. Very few companies have got this motion right – and ClickUp is among those.

For all of these reasons, I’m thrilled to share an in-depth look at how ClickUp blends PLG and sales-assist. If you want to learn more, follow Gaurav on LinkedIn or check out ClickUp (they’re hiring!).

---

Revenue as the North Star, embracing incrementality over attribution

I entered into growth from the perspective that revenue needs to be treated like a repeatable machine. You need to be experimenting all the time on how to make that machine better, and you need to be uncovering new ground to find opportunities. 

Revenue machines in B2B are changing very fast and are becoming more holistic than ever before. Especially in a PLG environment, there is no single revenue owner! 

There are multiple revenue owners. And then there are rocky overlaps – or, if we can operationalize these overlaps, there are compounding overlaps. How do you create an org where there is some healthy overlap, but where the goals are very clear and there is autonomy and clear accountability? And how do you build a culture with a company-first mentality? People can’t be optimizing for their own revenue areas in a silo.

You need a strong base for data and incrementality. In the B2B marketing world, I hear a lot about the word attribution. I don’t hear the word incrementality.

The consumer world moved from attribution to incrementality a long time ago, companies like Netflix pioneered that. You need to understand the incrementality of everything that you do. 

When I started at ClickUp as Chief Growth Officer, that was primarily growth product, acquisition, growth marketing, demand gen, and data. Now I am responsible for growth, acquisition, all of sales, post-sales, marketing, data, and systems. Across the teams, everyone has a KPI that directly ladders up to revenue – or where I can draw a causal inference against that. Sometimes causality could be a static number; sometimes it can be a little more anecdotal or a red, yellow, green indicator. But there has to be a strong relationship.

---

The five pillars of ClickUp’s PLGTM org

There are two main ways that you could divide up a PLG business. One is to take a segment by segment approach (Prosumer, SMB, MM, ENT). Another is to divide into self-serve and sales-assisted. 

Regardless of how you cut/operationalize your business, modern day revenue machines are very complicated and are truly a team sport. If not done well, it can create a ton of challenges. The traditional org structures fail here (pre vs. post sales, self-service vs. sales-assist, etc.). There are three fundamental issues:

Disenfranchised ownership

I have seen so many orgs get this wrong. Blending self-serve revenue into sales quota and then wondering why is their CAC so high. And then seeing self serve not as a revenue driver, but as a “pipeline driver” for sales. People buy seven figure mortgages self-serve these days, buying behaviors are changing very fast.

Or another one – leaning on success to do all the account management work, but then not giving them an expansion quota and assuming all expansion is driven by sales. I believe in shifting ownership to teams where they can own that KPI 70-80% of the time. Asking self-serve revenue to be blended into a sales quota doesn’t make sense because now there’s a portion of quota that salespeople would get relieved on, but have no influence over. 

Misplaced attribution

As we shift into a world where efficient growth is the mantra, unclear overlaps create avenues where cost gets hidden. For instance, if success is driving expansion, then are sales quotas adjusted to reflect that? In any modern SaaS business, your best self-serve revenue will move into sales-assist. Now your self-serve team will say that they’re not getting credit for the revenue they generated. 

Cost allocation

This finally brings us to the main concern, what part of my budget is driving up the CAC? In a direct to consumer/buyer model, self-service costs are primarily acquisition, growth, and product-led sales. In sales-assist, costs are more sales-led, pre-sales and post sales. How do you allocate marketing costs if you bring in self-service customers that sales-assist will also leverage? How do you justify investment in growth product that’s often very impactful to driving expansion and NDR across both self-serve and sales-assist customers? Modern day revenue machines also need modern day cost accounting measures to ensure that you are looking at your business in a connected fashion. 

For a bit, we saw a culture where leaders were trying to focus on absorbing revenue and posture – “who has the biggest revenue” vs. focusing on adding incremental ARR. 

Since then, we moved to a paradigm where leaders are responsible for specific business levers that together drive total revenue in the business. At the C-suite level, we moved away from a self-serve vs. sales-assist lens, to a segmented lens – acknowledging that all teams need to play well together to win the segment. By doing that, what happens is my commercial sales leader is genuinely interested about self-serve commercial as well as sales-assist. And my enterprise sales leader is interested in what's happening in self-serve because they know their book of business will come from the accounts that are self-serving today. My self-serve commercial team also cares about how the commercial sellers are doing; if close rates are high, they’ll be encouraged to send over even more top-of-the-funnel. 

There are still different leaders for self-serve and sales-assist, but that’s needed for the operational span of control. What I’ve found is that teams can still exist the way they would have existed, but their mandate needs to be mapped to where in the business they’re operating and what value they are driving. 

We’ve set up the business around five pillars:

1. Self-serve new logo. This is primarily a mandate for my growth acquisition teams. KPIs: number of new quality workspaces, predicted LTV of those workspaces, self-serve ARR. 

2. Self-serve net dollar retention (NDR). This is primarily the mandate for my growth product, lifecycle, and support teams. KPI = NDR.

3. Sales-assisted commercial bookings. This is primarily the mandate for my SMB sales team and within that go-to-market enablement and solution engineers. KPI = Bookings.

4. Sales-assisted enterprise bookings. This is where we have go-to-market teams and the ecosystem supporting them. KPI = Bookings.

5. Sales-assisted gross dollar retention (GDR). Sellers are responsible for expansion as well as new bookings and make a lot of their revenue through expansion dollars. But we have other groups like professional services that are responsible for sales-assist GDR. They also play a crucial role in delivering expansion, but are primarily measured on churn mitigation. KPI = GDR.

These teams are supported by two platform teams that don't play a specific position in a pillar and instead support all the pillars. In marketing, for example, this includes brand and platform product marketing. 

---

Cutting CAC payback in half by embracing the role of neutral observer and a culture of intellectual honesty

Because software businesses have been very high margin businesses by default, there hasn’t been the need to go find where the hidden buckets of inefficiency are. But in the current climate you have to do it.

I coach my leaders on two different things. For change management in your organization, we have to do right by people and ensure that the change management can go through well. But you have to almost separate yourself from being in that operator's shoes. 

Become an objective viewer of the situation and use that objective frame of mind to identify where the inefficiencies are. Then you can act on it. (And when you act on it, you should put your team leader hat on to figure out how to structure this change in a way that it goes through smoothly.) 

We break down our universe into discrete portfolio bets. Each of the five pillars get budgets based on how they are performing and their impact on the overall segment economics. It becomes a game of marginal efficiency and how do I continuously balance my portfolio so that the net of everything comes out lower than what it was before.

This starts with setting overall limits. Let’s say, for example, the overall limit is a 24 month CAC payback. I want to allocate dollars based on where I’m seeing growth, and translate those allocations into specific function-based goals. Then each team should be optimizing payback to make it more efficient. 

When we did this exercise across the five pillars, we realized we had untapped opportunities in many areas. For example we uncovered fast cash cycle opportunities with self-serve acquisition, our commercial motion proved to be much more efficient than we thought, the ROI on our support resources was very encouraging, and so on. We literally halved our payback there by doing the right things, and then we had more appropriate goals for our sales-assisted commercial and upmarket teams based on the investments we were making in them. 

---

Translating self-service interest into enterprise sales

If the customer could go self-serve, they should not have a sales interaction, right? You want to minimize the friction as much as possible.

With AI, there's going to be a band in-between where a lot of transactional, low average deal size (ADS) sales-assist will become AI assisted self-serve because there will be massive leverage from technology. Self-service just means that you don’t have to jump on a call or talk to a salesperson and instead can shop at your own pace. There will be experiences delivered to you at the right time when you want them and in the right way. 

The ultimate reason to talk to sales is because you are evaluating a bigger purchase. It’s been villainized that a customer should never have to talk to sales. While that's a good theoretical point of view, oftentimes for complex software products the customer doesn’t know what they don’t know. At a certain size of purchase, talking to someone helps you understand the context of your problem better, how this solution might fit in that context, and how you can deploy that solution better. 

People raise their hand to talk to sales when they want some of our enterprise features or when they're looking for a serious purchase. Sellers also have their book of business and they have to go build relationships into these accounts and do more value selling. That's where they create demand in the books of business.

PLG sales teams usually begin with a very inbound-first approach, a hand raiser-first approach, and often struggle to cross the chasm where they can’t move to a more relationship-driven sales motion.

If you don’t cross that chasm, you will never be able to grow a true sales-led growth pillar, which is ultimately needed if you have ambitions to get to a billion dollar ARR business. 

You need sellers who can create their own pipeline. They hunt down the buyer or the executive, convince them why this is the right solution, and sell a big deal. That’s very different from filling a 10 seat transactional order. You shift from transactional orders to fulfillment inbound to value-selling inbound and then ultimately to value-selling in a relationship-driven way. That journey is a very hard one. 

---

Using data to sell and redefining product-qualified leads 

You have to surface the best data you can about the account to the seller so they can act on it. It's not just product data. It's also account data, marketing data, who's attending events, and all that stuff.  I think we can do much better here, and we are getting better at it over time. 

And you want to build the most strategic and efficient playbooks as you can. When someone does this, you can actually use a phrase like X. Or you can talk about Y. You want to be as descriptive as you can. 

Our account scoring right now is primarily signal-based. There’s a bucket of data points coming in from product activity and another bucket that’s firmographic. The third thing we’re now adding are marketing signals – and I think marketing signals are where you can see how deep is the intent of this buyer even if this particular user doesn’t manifest through a product signal. 

Everyone thinks this is a deep exercise in data science. In reality, you can just come up with four or five dimensions. People are looking for precision that can tell you the exact 10 accounts that are going to buy. But even if I told you these 10 accounts will buy, you can only convert 30-40%. You can’t convert everyone, but you can eliminate the 98% or 99% or accounts that sellers should not focus on. 

MVP-based thinking is what’s needed so that you can operationalize the engine fast rather than waiting for that perfect, precise signal. People are waiting on perfect data rather than focusing on the “plays'' part of the equation. A product qualified lead (PQL) won’t reveal anything magic, but how do you take that data point to a bigger story and then run that play? There’s a strong operational machine you need to create around those signals that’s perhaps even more important than the signals themselves. 

Ultimately, what I see is that some sellers are really good at finding angles. A lot of what those sellers do is they just hustle. Everything in sales that can be broken down into a fixed set of patterns, machines will help with that. The ones who are very creative and who know how to hustle to generate alpha, those sellers will win. 

These sellers are great at talking to people in an account to understand the context of the problem, figure out the company goal, and build an argument for why something like ClickUp should exist. Through that effort – or through outbound outreach, email, or LinkedIn – they’re able to map to the executive buyer and convince them why ClickUp should be used in a way that’s bigger than how it’s used right now.

The ones who are really good at it are extremely surgical. They build an account plan where they map the needs and the buying panel in an account, then work each one of them. That’s how they land us six-figure or seven-figure deals. You land iconic deals by being super intentional in your books and crafting a very bespoke business case to get the attention of the buyer. 

An org that can figure out how to propagate that creativity throughout the entire sales org, that org will win. 

---

Embracing usage-based paywalls over feature gates

Where ClickUp made a very solid choice – and the credit goes to Zeb, our founder – is that we use a lot of usage-based paywalls. We always provide some amount of usage to almost every segment, and that way everyone can experience a majority of the paid product in a limited way based on the pricing plan they are in. Based on what the customer is using, they can move up a plan if they think they need more. We rarely do hard paywalls and even the hard paywalls we do have, we’re converting them into usage-based paywalls.

ClickUp’s vision from day one has been the everything app for all of your work. We’ve always focused on giving more and more product features to our customers including project management, docs, clips, whiteboards, automations, and more. All of that adds compounding value and so for us it’s about letting users try everything. 

There are some features that are mostly enterprise-oriented, which we keep within our enterprise plan. For example, SSO-based login, data security, data localization, or other things that have a cost of deployment. You want to keep those in the enterprise tier because if a company is looking for it, they’re most likely a very serious buyer and will have the propensity to pay for those features. 

---

Closing thoughts 

Don’t underestimate the impact of a human touch. The problem is that a human touch delivered randomly is very hard to optimize for. But if you can deliver it in an opinionated and systematic fashion, it can have massive multiplier effects. 

Combining growth teams with GTM teams can unlock value that’s never existed before. We paired our lifecycle team with our customer success team and the result was massive – I’ve never seen such a lift in NDR – because we’ve been able to augment customer operations with growth marketing-type thinking. We’re trying similar things with XDRs and demand gen, as well as commercial sellers with systems and technology. 

It’s hard to innovate if your livelihood is on the line. People should have some variable pay, but it should be about giving them upside instead of reprimanding people through pay. Using pay as a way to control quality creates bad incentive structures and doesn’t leave room for innovation.

Using compensation as the only way to hold accountability is an inhumane way to run a business. This is something I am personally motivated to change in the years to come, a lot of the SaaS playbooks will be re-written and I hope to preserve the upside for top performers in GTM, while still giving them room to be innovative and take risks.