Factors.AI is an end-to-end marketing analytics platform that integrates across data silos to deliver focused AI-fueled actionable insights. We provide answers to critical questions on marketing influenced revenue, factors affecting lead generation, paid ads ROI and multi-touch attribution through a “search-driven interface”, automated insights and analysis template library.
Praveen: Monish, we could start with you giving a quick overview of yourself and your journey.
Monish: First of all, I appreciate you bringing me here. Hello everyone, I hope everyone is doing well. I did Industrial Engineering as my undergraduate. I liked it from the beginning because it was more of a liberal art than engineering. So it used to tell you how you can add more value and reduce waste. That is the fundamental thing on which industrial engineering is based and went onto be a significant influence on my future work. And then, I started my career as a Supply Chain Consultant with HP. In 2008, I joined the Digital marketing COE at HP, and initially, it was more tactical landing page creations and brand compliance and all that. Then eventually moved on to much more sophisticated things.
Three years later in Digital Marketing COE, marketing technology was taking a new shape. And data analytics was becoming more relevant in the organisation. And I joined the marketing operations organization as Senior Manager of Marketing Effectiveness. I was responsible for setting the entire marketing data infrastructure along with the BI team.
HP split happened, and I moved to Red Hat, which was another unique experience where the organization was selling a product that was available for free in the market. Linux is available for free in the market, and everything that they sell is available for free in the community. But it was a $4 billion organisation. IBM acquired it at $34 billion, which was a huge success story. But it was a new thing for me to learn as well. And it was also exposure for me into the SaaS business, to begin with. Post RedHat, I joined Arun Pattabhiraman’s team at Freshworks as Director, Marketing Operations.
Praveen: What does marketing operations at Freshworks mean, and how does it play within the larger marketing function at Freshworks? What are the various sub-teams with a deep dive on marketing operations?
Monish: My mentor from Redhat used to quote by Louis Pasteur, and in the time of the pandemic, he becomes even more relevant. He used to say, "Chance always favours the prepared mind". So what we are doing is we are giving every marketer a chance to be more efficient and to be more successful in this journey. And that is what we essentially are trying to do. So if you look at marketing operations organizations, we have 1.) Measurement and analytics, which essentially is going to be responsible for creating all the right metrics- leading and lagging metrics, measuring every team's performance and also running analytics which helps us scale on the longer term. Then 2.) The data infrastructure management is the team that works very closely with the business intelligence team in setting up the right marketing data infrastructure. As and when we bring in more marketing technologies onboard, there are two more important things that we have to do. One is to lay the workflows so that the data pipelines and the integrations are correct. And second, we can ingest as much data from that platform as possible. Then 3.) Planning and budgeting team. We look at the entire growth and marketing budgeting and are responsible for building the annual operating plan for growth and marketing Next team- 4.) Marketing technology stack- is very self-explanatory but what we have done intentionally is not to buy technology for the sake of it. We make sure that we are buying technology that is fitting a piece of the puzzle that is missing right now and is making us better over a period of time. Then 5.) Best in class processes- this team defines processes and workflows so that we build workflows for whatever technology setup is there which enables better collaboration between sales and marketing teams. And also help us provide the right data model that we'll be able to create in our data infrastructure. Then next unit 6.) Agile marketing is essentially marketing enablement. For every function and every region, we have a program manager for marketing operations who's like me representing that particular function. That way, we can have better controls in terms of how we do everything, whether it's setting up campaigns or using a marketing piece of technology, or using audiences. Finally, 7.) we have added orchestrating ABM and audiences as part of the charter. For the scale of our organization where ABM is not going to be restricted to one-to-one, which is focusing on one account because our average deal size is not the right deal size for one-on-one ABM. So we want to do ABM at scale. And it is going to be driven by an audience-first approach. So this is how the current marketing operations organization is.
Praveen: So when you say best in class processes and agile marketing, I'm presuming that is more program management to speak- e.g., are we executing like in two weeks. Is that the right way to understand agile marketing?
Monish: Yeah. I'll just add one point to it is that we want to be a strategic partner with every function. So if field marketing is running a campaign, we don't work with them on running a campaign during the quarter. We work with them to understand what your pipeline target for the quarter is? How are you going to get to that target if you run these campaigns and if you target these audiences? And once they run those campaigns, we put a hawkeye of a daily view of whether they are generating the right CPLs, whether the leads are getting generated, whether the score of those leads is right, whether at an account level we are generating the interactions that we need to. There are multiple channels through which we'll be making investments. So understanding which channel is generating the right bang for the buck for you and which channel is not. Then should we de-focused or de-prioritize that channel and double down on our investment in the channel which is generating good results. So that is where we are bringing in this closed-loop between planning and execution. We are at a stage where planning feeds into execution and execution feeds into planning. This is what is coming in 2021. We spent most of 2020 to build the right framework and infrastructure so that we can build the right team of Agile marketers in 2021.
Praveen: Got it. So then the agile marketing team works on top of the planning, budgeting, measurement, and reporting work to bring context?
Monish: They are the owner of that function and help them in providing everything useful for them to run better marketing.
Praveen: Got it. And the best in class process team, is it also more of guidelines, systems, and processes?
Monish: It is, and it has a different connotation as well. It will cover your touch governance policies and your customer life cycle status. It will also cover how we treat the contact life cycle if we have like 3 million records, how do we scalably work at a point that we don't see that there are a lot of zombie contacts in our database? Also, how do we do retargeting? And if they are not engaging with this with us, how do we archive them. Because if they are not engaging with us through emails, we need to either reactivate them or those emails have become invalid for us. The second thing is the workflows that we need to define between sales and marketing. We are also in the process of completely transforming our focus from lead and just getting rid of that terminology completely. And we want to focus on the account. We will move from the traditional demand waterfall to a more new decision demand waterfall which is based on the demand unit and everyone is getting measured on the pipeline. We now have a multi-touch attribution model where we can attribute paid campaign impressions to the pipeline. So what we want to do is put the workflows in place which are reflective of how B2B organisations engage with us, whether it is SMB or Enterprise.
Praveen: I think just for the benefit of everyone, could you next spend maybe one or two minutes on the twin-engine model for Freshworks? The kind of the two extreme classes of customers that you have so that people understand that they're not like one uniform set.
Monish: Yeah. We have a very assembly line kind of a model for SMB. When we talk about the twin engine model of Freshworks, we have a low-touch high-velocity business in the form of SMB and we have a high-touch low-velocity business in the MidMarket, Enterprise business. The SMB engine is predominantly driven by inbound engagement with the customers and for midmarket enterprise, being a more buying group based decision making, has a multi channel approach from both marketing and sales.
Since our products are built on the foundation of ease of use, we are also a firm believer in product led growth where the growth of our products, especially in midmarket and enterprise segment is not only driven by decision makers (CXO’s) but also by influencers who use our products and then champion its implementation in the organization.
Praveen: Do you measure different metrics for tracking the effectiveness of marketing for these two buckets? Do you have a separate set of reporting dashboards or slides for each of these?
Monish: Yeah, we have a separate set of reporting dashboards because in the case of SMB, it will be useless if we measure pipeline because it is a high-velocity business. So there is no point in having leading metrics like pipeline there. In that case, the leading metrics become your website visitors, website conversions, your sales conversions, how they are converting, what are the different employee sizes within SMB who are coming in? How we can go more granular to understand our SMB segment more in details by further splitting them into sub segments and creating different cohorts based on their buying parameters (average deal size, conversion rate, etc). Because each of these sub-segments is going to give you different average deal sizes and we are looking at which sub-segments bring in most of our leads and are they bringing proportional revenue or not? For non-SMB, it's more traditional B2B buying group-based and persona-based. And then you need to have a marketing source and marketing touch for the pipeline and how marketing is contributing at different stages. So that is why you need to have different metrics for them because the business models are very different.
Praveen: Is there any sort of customer marketing like sending out case studies, email, and whitepapers, etc.?
Monish: We have an outstanding team led by Yasasree who leads customer marketing and it's growing exponentially. The biggest problem in the SaaS business is that if it is a leaky bucket then you can add more and more but if users are leaving then your net numbers will always be very low. Overall from a Growth & Marketing standpoint we have a phenomenal vision driven by Arun Pattabhiraman, who set up functions like customer marketing and segment marketing which are significantly contributing to our customer growth and retention.
Praveen: How do you see marketing influenced revenue?
Monish: How we see marketing sourced and marketing influenced pipeline is some sort of a retrospective view of whether the contact which is associated with an opportunity, had how many campaigns touch before it became an opportunity. So that we not only look at a larger window of the last touch before the person becomes the opportunity but also the campaign touches during the overall deal cycle before it becomes an opportunity. And then we run a machine learning-driven multi-touch attribution model which gives weightage based on the removal effect of that particular touch.
Praveen: Got it. So I have two more questions. The first question is- From our conversations, the last 12 months have been like a high-velocity sprint, and you've been heavily enhancing the capabilities of the marketing ops function. So if you could reflect on that and tell three most meaningful things- which have entirely changed the way the team operates and focusing more on the analytics and reporting function as that's probably most relevant for us. Also, what did you see when you came in? What were the changes that you made and so forth?
Monish: I think there was one thing that we changed which is extremely important- changing the narrative and changing the conversation. We are creating a strategy and narrative around Ideal customer profile and account based strategy which is different from marketing focusing on leads and handing it off to sales. We focus on optimizing our media spends and looking at a diversified view of spending our media dollars by building a cohesive audience strategy. All this brings in better alignment between sales and marketing and also elevates the narrative towards business outcomes.
The other core thing is that we are now at a point where within five minutes, 80% of our leads are getting assigned to the right sales rep. From the time a user enters into our system to the time they get assigned, we do enrichment, territory assignment, rules of engagement, whether it is part of a named account or a non-named account, and lead to account matching done within five minutes of the SDR or sales rep (depending on rules of engagement) getting assigned, and that is true for 80 % of our leads. So it increases your lead velocity from days to 5 minutes. That is a significant impact of our work.
Lastly, when we started working on our annual operating plan (AOP), while the guidance used to come from top management, we are able to provide strong data background and logic to the AOP with a bottoms-up model of seeing what is our total addressable market (TAM) by-product, how much conversions we can see this year by region and each persona level and then arrive at what is our overall revenue potential? This leads to what is a realistic revenue we can do next year and what is the pipeline needed for the same? This was all done from a bottoms-up level and gave great confidence to everyone in executing the AOP.
That is very essential and that is how as a data-driven Marketing Ops team, you can change the conversation and narrative in the way that you help in building what the right path for a company to grow is.
Praveen: Understood. And on the data and analytics front, I believe one of the investments that you guys made was creating a new data infrastructure for storing all these marketing interactions and being able to drill down into each opportunity to say what all happened to that opportunity from the time that they came into the system. So what was that like? Was it an expensive sort of an effort?
Monish: Freshworks already had an enterprise data lake. We partner very closely with a business intelligence team and they are bringing a company wide enterprise data Warehouse which will enable democratizing of the data for the business users like marketing. We used that to build our own federated data mart and called it “marketing Intelligence cloud”. For visualization there was also the company wide BI platform that was brought in which we leveraged to generate metrics and insights reporting. We have added an analytics layer on top of the data mart which helps us in building predictive analytics models like multi touch attribution, that we spoke about earlier and also other forecasting models which helps us in making better decisions.
Praveen: Understood. I would want to just leave it open for the rest of the team.
Audience Question: From the data sciences side, what gaps do you see which people thought or promised they would fill, but did not or could not.
Monish: I think one of the things we always made sure of is that we don't build models just for the sake of building models. So we need to apply some logic to that before we see how we can use that model. The most significant gap that I see in data science is that first of all if you have a lesser amount of data, it's just going to give you terrible results. Data integration from channel to CRM will provide you with all the data that you need. What we were doing was that we were doing both these things in parallel- building data infrastructure and data modelling. The biggest problem that we used to face is that we used to have historical data but the data was not in the shape of how we wanted it to be. And the technology pipelines that we were building were not giving us an adequate amount of historical data because they were also getting built in parallel. So the biggest problem that I saw with data science is that we should not be building teams until and unless we are ready with the right amount of data, the right data structures, and the right kind of infrastructure from a technology standpoint. So that is why we didn't build a huge data science team this year. But now with the whole technology getting integrated very well and the data getting ingested for the last 12 to 18 months, we are seeing a really good performance of the models.
Praveen: When you are hiring a data science team, what are the top 3 things you look at from a business problem statement or marketing point of view?
Monish: For data scientists that I need to hire for marketing, I need to have a person who is coming from a liberal arts background like statistics or econometrics. I consider statistics as a liberal art. I need those people rather than people who are engineers and who just know how to run packages in R and Python because that's very easy these days. If you want to run a random forest right now you just have to ingest data, and with little understanding of R or Python, you’ll be able to run it.
Praveen: Within marketing, what would be those three problems that you would have identified as the most important ones for Freshworks?
Monish: Three things. The first is how do we optimize our media spending? Understanding how we are spending on paid, owned, and earned media? Paid is everything that you spend on external channels, owned is your website, and earned is your social media activity. How do you get all these user touchpoints and events together? We need to pull all the audiences from all these channels together and then optimize your media spends based on the audiences and this is going to be the most important and critical thing that we'll be doing next year.
The second is going to be around sales analytics and capacity planning. What are the leading metrics that we need to give to SDRs? How do we improve SDR productivity? And we need to look at deal and revenue intelligence using conversations to provide intelligence around how we are going to understand whether the deal is going to close or not.
The third thing is going to be customer and TAM analytics which is looking at how can we expand from the existing customers, how can we make churn predictions so that we can reduce churn, and how can we look at overall TAM and do prospecting around understanding which prospects have higher propensity to buy.
Audience Question: About twin model engines, considering a three years time frame, how was the percentage of money and efforts spent on SMB as compared to enterprises because they seem to track different metrics?
Monish: Though the metrics were tracked separately, the North star metrics were always revenue and pipeline. Midmarket and Enterprise being high touch need more investment. For SMB, we are dependent on paid media but with the right team structure being created by Arun, we are seeing proof points of how non paid can generate better results. It takes time but it gradually becomes a humming well oiled engine.
Audience Question: So what touchpoints might be present for SMB and enterprise segments considering the marketing and sales as a single function?
Monish: Whether its SMB or midmarket or Enterprise, any campaign touchpoint where there’s an explicit intent shown by a prospect or customer to engage with them is considered a campaign touchpoint. This essentially means all the low intent touchpoints like email open clicks etc are not considered a touchpoint. We need a PII (email id and contact info) submission in order to consider that touch a valid touch for attribution.
Praveen: Thanks, Monish. From a sales and marketing alignment perspective, what have you guys done specifically which has proved to be effective? Are there handoff points, metrics, things that you have defined that have made that relationship easier and smoother?
Monish: it's not essential whether the relationship is smoother. The important thing is whether we have the same metrics that we are focused on. Eventually, there will be friction and that’s good as it pushes both the teams to drive better outcomes. But we are trying to converge to the same metrics now. The good thing is Freshworks, where sales and marketing have the same metrics, which are Revenue and Pipeline. Not like other organizations where you have a Marketing organization, and you have a Sales organisation and Marketing is always talking about leads, and MQLs and Sales are still talking about pipelines. So for me, the important thing is not around friction. Friction or better way to put it is that debate and discussion should still be there. The collaboration should be driven when metrics are going to be the same thing. So the only thing that you can do in driving better collaboration is taking the subjectivity out and how you take the subjectivity out is just when you put data in front of everyone which is very transparent and shows everyone what is happening. So that is what we have been doing in the last 12 months that we slowly and steadily take the subjectivity out of the conversation and have just the same metrics getting represented across everyone and the source of that metrics is also the same.
Srikrishna: Thanks a lot, Monish. Thanks for your time with the team. And it's a lot of useful insights from the business side for us to pick up from here. And it'll also put a lot more meaning into our efforts from the engineering side and product building side as well.
Monish: Most welcome, Sri. Thank you, guys for having me. Bye-bye.