Roll up roll up... If you're looking to re-design or implement your sales and marketing funnel, then you may well want to read all about my interview with Denis Malkov, Director of Rev Ops at high growth PandaDoc. Denis' insights were laser-like and the structure with which he approached the various topics rock-steady. After the usual WebEx connection palaver, the interview began...
Rory Brown (RB): Can you tell us about your career and how you got into rev ops?
Denis Malkov (DM): I joined PandaDoc about two years ago in a business operations capacity, working on helping move the company towards a repeatable, scalable and predictable revenue model. Early on, we identified that there was a big gap in alignment between the revenue teams, particularly marketing, sales and customer success, and some support as well. At the time, we had a really bloated tech stack. We had different teams running in different directions and there wasn’t much of a feedback loop. We also lacked a process for handoff, documentation and good reporting. And we didn’t have a funnel. We had no idea. We get about 1 million visitors to our website every month and we had no idea how those visitors cascaded down to closed won business and beyond into expansion, renewals, and referrals.
The rev ops team concept was born out of this need, not just to align these revenue functions, but to instrument the funnel for the company and start to maximize conversion rates and increase velocity between stages.
I started building the team in early 2018. I hired a Marketing Operations Manager, Sales Operations Manager, Customer Success Operations Manager and we also grew a Sales Enablement Manager into a more holistic revenue enablement manager role. She leads training and onboarding programmes for all of our revenue teams.
Prior to PandaDoc, I was in finance. I worked at GE Capital for the first part of my career in financial planning and analysis and in business development. Then I got my first taste of startups with a company called Adroll, a re-targeting company turned ABM company. I then worked for a number of smaller startups along the way before PandaDoc and I found each other.
RB: You are the first person I’ve interviewed who came into a company with no funnel. Congratulations! I’d love to hear more about that.
DM: We had a two-stage funnel: The Top and the Bottom. That was it!
RB: Where did you start?
DM: We started with a buyer’s journey. Mark Roberge at HubSpot often talks about taking a customer-centric view of not just your funnel, but how you present your product and how you present your business externally. So, we started by trying to understand more about our buyer, our ideal customer profile and what sorts of personas were coming into PandaDoc and finding success with our product. And that buyer journey mapping is really the top layer of what we consider to be our funnel. The journey from awareness to consideration to a decision to value creation is crucial, because if we don’t understand how our buyer is becoming aware of and buying PandaDoc, then we can’t possibly hope to implement stages to help them through that buyer journey.
So that’s where we started. Then underneath that, we built a layer of PandaDoc stages, which is our true funnel. We have people at the top, visitors, traffic. These can be unidentified, or we might just have a few details. Afterward, if they do something on our website or they download some content, that is another stage. Then after that, we might decide they are an ICP or they are the persona that we want to target because of the buyer journey. That is another stage where we pass them to a BDR.
And then the funnel keeps going. Each stage has a very clear point of conversion that is measurable and can be architected in our system. So that second layer under the buyer journey is where our funnel sits. And then underneath that funnel – a third layer – are the systems and the activities behind it.
RB: Let’s take that first point, understanding an ICP persona. Where did you start there? How did you gather that information and ensure that it was reliable enough as well?
DM: In a number of different ways but the two main categories were quantitative and qualitative research. On the quantitative side, we did a deep, deep dive into our own data to understand the different segments of people using our product. Because we have a very horizontal product, there are lots of different use cases for lots of different industries. So, we started by looking at the mix of accounts from a firmographic standpoint, primarily company size and industry and we found that PandaDoc is really its own ICP, which was really nice to find out because that means that companies like PandaDoc use PandaDoc – high velocity, business to business SAAS companies.
This meant we were able to build an ICP profile looking at ourselves and understanding ourselves, because companies like PandaDoc were the most likely to retain, had the highest average revenue per account, had the highest usage within our product and were using the things that are part of the PandaDoc value statement in our buyer journey.
RB: Moving onto the second layer. I have spoken to a lot of people about the handover of leads from marketing to sales and I’m really keen to understand how you guys have been doing that.
DM: Our SDRs sit on our marketing team. Marketing drives both demand generation at the top of the funnel, and then pipeline creation for our sales team. An SDR’s single most important metric is what we call Stage Zero opportunity pipeline.
Before we did it that way, pipeline generation from inbound was a shared responsibility between SDR and AE but that created a lot of fuzziness in our funnel. Also, it was creating a two-part role for AE’s as they were prospecting and closing, and it really didn’t orient them to the right metric - closed won business. So, we split that out and we said: Stage zero pipeline creation is the SDR and marketing teams’ full responsibility.
And to be clear, the responsibility doesn’t stop there. Because we have a funnel instrumented and we have feedback loops established, we can see if something is closed lost in stage three or stage four that the demand generation behind that pipeline was a little bit off. Perhaps they came in for an eSignature use case and they made it through the qualification criteria with the SDR and they made it into the AE’s hands, but the use case was never right from the start.
Now that we have the funnel and we have the feedback loops working, we can give that feedback to demand gen and say, “We really need to be bidding less on eSignature keywords because it doesn’t close well for our business.”
To get back to your question, the marketing team owns pipeline generation. They take a lead and run it through qualification criteria which is similar to the age-old BANT – Budget, Authority, Need and Timeline, but really focusing on need as impact i.e. what the impact is going to be to the business. Again, this is getting back to the buyer’s journey and customer-centricity. So, they qualify it and then convert the lead into an opp. Then the opp is passed to an AE to progress down the funnel.
RB: Fantastic. In that circumstance, what are the key metrics that marketing is working towards?
DM: They own pipeline generation as their top-line KPI. But then we’re looking at closed won rates. I don’t really care if you generate $1M of pipeline a month if none of that pipeline closes. So, in looking at the marketing end-stage and sales part of the funnel we can see that marketing created all of this pipeline and then it got passed to AE but then none of it closed. This indicates that we need to go back to the SDR team and coach on qualification.
RB: You mentioned two key metrics at the beginning - conversion rates and velocity. Why did you choose those two key areas and are there other metrics that you throw in there as well?
DM: The two metrics – conversion rates and velocity – are what drive our North Star metrics for the business, which is Number of Customers and Average Revenue Per Account. If we increase the number of ICP customers, we increase average revenue per account. Those are the two main goalposts that we have set company objectives around and flow down into everybody’s department and individual OKRs.
RB: Awesome. When you sit down and analyse your sales cycle to try to reduce it, where do you start? How does that process work for you?
DM: PandaDoc’s average cycle is about 30 – 45 days. And we are focused on the B2B, SaaS, high velocity sector which typically has company sizes of 11 – 500.
When looking at accelerating the sales cycle we do a couple of things. Firstly, we make sure that we have really tight qualification at the SDR level, the pipeline creation stage. We are not just looking at a bunch of unqualified meetings and sending it to AE who then get stuck in Stage 1 forever because the qualification wasn’t tight enough.
Secondly, we have SLAs where the SDRs are partially incentivised on Stage 0 to Stage 1 progression. Stage 0 is basically Qualified Meeting Booked and Stage 1 is Sales Accepted. If we fall below a certain percentage threshold of Sales Accepted, the SDR’s compensation is impacted. This means we ensure we are passing qualified opportunities, not just as many opportunities as possible.
Third, we have a tight forecasting process in place. We use a tool called Datahug to submit forecasts on a weekly basis, which gives us a top-level view per sales team, whether you’re SMB, mid-market, upmarket or account management. It gives us a per-team view of forecasts for the current month and quarter in terms of pipeline, best case and commits, which we then use to coach in one-on-ones with sales reps. We will look at Datahug forecasting and go deal by deal to understand what the next steps are, what the level of activity is at each account, and whether this is realistically closing in the next 30 – 45 days. If it is not, we will aggressively close lost things in favour of new business that has a higher propensity to close faster because we want to keep our velocity really high.
We are a 95% plus inbound driven business. Because we are getting so much traffic, we have a huge amount of pipeline creation. That means we have to be picky about the pipeline that each AE holds at any given time.
RB: Where do you start when it comes to a conversion rate between a particular stage? How do you improve that conversion rate?
DM: The sales conversation has to be driven by pain points. This comes back to our buyer’s journey and in mapping and understanding our buying process. What are the pain points that bring you into the product in the first place? First, we have an SDR qualifying those pain points. Then, we have sales re-qualifying those pain points. And then even once we are closed won and in the implementation phase, we have customer success revisiting those pain points and setting the product up to make sure we are tackling those paint points and showing improvement to our customers.
So in the sales stage, keeping things moving and accelerating is comes down to drawing those pain points out as early on in the conversation as possible. And then showing explicitly how the product can help the customer solve those pains.
RB: Perfect. You mentioned the feedback loop, which sounds like something relatively new given the lay of the land before all of this was set up. First of all, what was the idea behind that? And how do you set up an effective feedback loop so that it is working and adds value?
DM: I think the problem statement, or the basis behind setting up a feedback loop, is it is an inefficient system to have marketing process leads and create pipeline and hand that off to AEs and hope for the best. It is really inefficient for an AE to close won an opportunity, and to hand that off to CS and hope for the best. What that creates is disenfranchisement between teams You have sales getting upset with marketing because they are not passing high-quality opps. You have customer success getting upset with sales because they are over-selling or over-promising onboarding experiences where we shouldn’t be in the first place.
That is the problem statement behind the feedback loop. Getting back to the funnel, what the feedback loop looks like is each stage has a clear gate behind it, so for us going from Marketing Qualified Lead to Sales Qualified Lead is a positive response. If that positive response doesn’t happen, and suddenly we see MQL to SQL drop off precipitously, we know that our message is broken, and our cadences are failing in some way. We know exactly what team to go back to and what process to tweak in order to get that part of the funnel re-established.
If SDRs are creating a bunch of pipeline for Account Executives that are not closing, we have very discrete closed lost reasons to look at those reasons and understand why things are dropping out. And then give feedback to marketing and the SDR team to start qualifying more efficiently.
RB: When you have a closed lost reason in Salesforce, typically one of your AEs has lost a deal. There is an element of pride that comes into that scenario whereby the reason that they disclose might not necessarily be totally indicative of what actually happened. How have you found that situation? How do you breed a more honest culture to avoid that scenario?
DM: Transparency, honestly. All closed lost opportunities are exposed in Slack to a public Slack channel. That doesn’t just include sales management and revenue operations but also product managers and marketing managers. They see a deal amount, who closed lost the opportunity and then the closed lost reason behind it. And whether it is timing related or feature related or the opportunity just ‘went dark’, that’s all available to every single person at the company. So, we can see in real time why we are losing deals and help each other understand those customer pain points better, and then coach each other. Or maybe we just need to give clear feedback to the product team so that they can build things into their development cycle if we are seeing a bunch of closed lost opportunities for feature-related reasons.
RB: Yes. I really like that idea. That is the first time I’ve heard of it. Presuming the fact the AE knows everyone is going to be reading it, they have to be honest.
DM: Yes. They know everybody is reading it and they know the company takes it seriously, especially our product management team. They are constantly trying to build a product that is valuable for our customer base. That sounds really straightforward and obvious, but not a lot of SaaS companies do that. They just build what they think is best, without really listening to their customers.
RB: Brilliant. How do you find the balance between product feedback based on closing a sale and product feedback based on customer retention?
DM: Yes, this is a mechanism that we have put a lot of rigor around over the last quarter. There is a rev ops driven initiative as well, through revenue enablement, which is a revenue team and product feedback loop. This is looking outside of our funnel, instead of looking at our product development cycle. Product and engineers are very , and they should be. So, revenue teams need to give them a very straightforward way to sort through data, to understand prospect and customer pain points so they can build the product around it. For sales, we have closed lost reasons, which works really well. We can look at the product feature related reasons on a monthly basis, and draw out: We need redlining in our product, or we need single sign-on because a lot of customers at this value are asking for it.
But then looking higher up in the funnel to marketing, we have built a similar mechanism of “Are we disqualifying a bunch of leads because we don’t have this built into the product yet or because it is a partial use case.”
Similarly, on the customer success side if there is an expansion opportunity or we were unable to retain an account – if an account changed – we have a similar feedback mechanism set up in our CRM. And then every month we sit down with our product management leaders and we draw patterns out of this data at the marketing, sales and customer success stage so we are not over-indexing on any one part of the funnel. We might say, “OK, we are seeing patterns in not being able to qualify, close or retain business that is related to this specific part of the product. So that is obviously valuable to all parts of the funnel.” Not just “can we acquire more people”, or “can we retain more people”.
RB: Great. Moving onto the marketing to SDR handoff. I’m guessing if you are a Mark Roberge fan you have got SLAs or something similar in place? How does that look for you?
DM: There are a couple of components of the marketing SDR SLAs that we talk about. First, we lean pretty heavily on automation for our non-ICP cohorts. So again, because we have very high top of funnel traffic, we need to only serve customers that we think will be valuable but not necessarily put a human in front of them.
When I say ‘put a human in front of them’ I mean white glove, one-on-one SDR to prospect interactions from day one. We have cadences built out for smaller, less ICP-centric businesses that are fully automated. We will send them email drips and content and once they respond, we will put them into the right person’s hands. But we are not going to be writing customized emails, reaching out to them on LinkedIn or sending them direct mail from day one.
That one component is again looking at whether they are ICPs and managing them appropriately.
Secondly, for those that we do put an SDR in front of from day one, the SLAs that we have are oriented around call volume, email volume and time to the first touch. Typically speaking, an SDR will perform an average of 80 – 100 activities per day. And that volume is high because, again, we are very inbound driven. They will be making somewhere between 40 – 50 calls and sending about 40 – 50 emails on any given day, which is the bottom-up activity metric. Then for response time on chat (we use drift at the top of our funnel on our website), we have a sub-1-minute SLA because those are high-intent, highly interested leads that are coming in and they want to talk to somebody now.
RB: Do you have one of those ‘Don’t be on it’ dashboards?
DM: We do, yes. We send a daily SDR production dashboard - we call it our Bill Walsh dashboard - which is all of our SDR plays. It shows any leads in an SDR’s queue that haven’t been called or emailed yet, with that SLA as a target time. It shows any low activities that we particularly identify as high target markets. And we have a number of other reports that get emailed out to SDR manager, sales manager, revenue operations, CRO, and a bunch of other people on a daily basis.
RB: Is that named after Bill Walsh the football coach?
DM: Yes, exactly. Bill Walsh was a head coach for the San Francisco 49ers. He popularised the Westcoast offense, which emphasized passing plays.
They would build out a passing playbook, and this is something that we are trying to do for all of our revenue teams.
RB: Brilliant. One topic I’m keen to explore with everyone is how we measure success in sales ops or in rev ops. What does measuring success look like for you and your team?
DM: Success for rev ops is maximising conversion and velocity, or funnel. We are not quota-based, obviously. Measurement for any overhead team like ours is a little more difficult and a little more project-based but we consider ourselves the stewards of the funnel. We do target ourselves to a certain lead to MQL conversion rate and then marketing qualified to opportunity. Between opportunity and closed won, obviously, we have less control as that is more sales. Then moving beyond closed won into retention, that is more customer success.
But the main part of the marketing to sales funnel is where rev ops gets its discrete KPIs from.
RB: Does the business give you heavy KPIs or are they happy as long as they see improvements?
DM: We have benchmarks. MQL to opportunity: 20% plus, which a lot of companies would consider pretty high, but again for a heavily inbound and a high intent driven business, it’s our bread and butter.
And 20% plus for MQL to opportunity. And then for marketing suspected, which is the first time that we are scoring these leads and passing them into an SDR’s hands, that shows that our lead scoring engine is operating correctly and efficiently and that we are sending the right leads to our SDRs. We typically look at it like 80% – 90% of MSL to MQL.
RB: Lead scoring is another interesting topic and I’ve heard arguments for and against it. Perhaps you could share why you chose lead scoring and how you put it together so that you are not missing people who have a low score and are high intent?
DM: Yes, we absolutely had to do it because we have 1 million website visitors every month and somewhere between 15,000 to 20,000 trials every single month. We are currently at six SDRs. Asking them to process every single one of those trials/demo requests/form fills would be an impossible and very inefficient task.
We built out a system internally. We have a data science team who are magnificent and they’ve created something that in my opinion completely trounces any software that I’ve come across in the business that claims to do the same thing. They built a behaviour and fit-based model, with fit being demographic, technographic, firmographic criteria -- company size, industry, and then persona.
Behaviour here means what they are doing on our website, what content they are interacting with and what they are doing inside the application during their free trial.
Based on those two categories or criteria of a lead, we will give it a score that is A1 through D4, A’s being activity or behavior and the 1’s through 4’s being the fit.
You could be a high fit but low activity or you could be high activity and low fit. And we will decide what to do based on that matrix that we have built out.
RB: So it is a multi-dimensional score. I like that, it gives them more context.
DM: Right, it is not a discrete number. It is a letter and number combination that gives them a better idea of: Maybe this is a high intent but non-ICP lead that I probably don’t need to customize an email and do a bunch of calls. Versus: This is somebody who is an extremely good fit for our business, is right in our persona or ICP wheelhouse, they just haven’t interacted with our content much yet. So I am going to throw everything that I have at this lead and truly try to understand what content I should be serving up to them.
Want to get more insights from sales ops leaders? Check out our other posts in the sales ops interview series.
We recognise the growing importance of sales operations. No longer seen as the function that provides spreadsheets, sales operations is integral to building a repeatable, scalable sales machine.
That's why we built Kluster. We make analytics and forecasting systems for you so you can spend time doing what you do best: uncovering trends and delivering growth defining insights.
Kluster gives you total visibility into the effectiveness of your sales machine and helps you generate credible forecasts to revenue leaders and the board.
Discover why Kluster, the leading revenue analytics and forecasting platform for B2B SaaS companies, has received multiple awards from G2 in their 2022 Winter Report. Kluster is the number one platform worldwide for Opportunity Scoring, Risk Analysis, and Live Forecasting. Learn more about the platform's features, awards, and fast integration times on our website.
Pravesh Mistry, Chief Revenue Officer at Truework, believes a strong level of clarity is what makes a good culture!