Siddharth Deswal: Making Marketing Operations & Analytics Simple For A B2B Saas Startup

Moderator: Siddharth, could we start with a brief background about yourself, and your journey?

Siddharth: Currently, I'm the Director of Marketing at Fyle, the expense management software. I started my post-MBA career in 2012 with Wingify. I was the first marketer and the 16th person there. I saw the growth of Wingify from $1.5Mn to 15 million ARR. I quit in 2017, after which I consulted for 2.5 years with various SaaS companies across India on different marketing projects. I have 8.5 years of marketing experience.

Moderator: Got it. So, a relatively broad exposure across SaaS companies for ~8.5 years, and now in marketing at Fyle. Could you elaborate more on Fyle as an expense management solution?

Siddharth: So an expense payment solution is needed when you have employees who are traveling or making expenses to carry out business activities. Most expenses used to happen around travel, stay, and transportation. Now, people need a way to capture the expense they've made on behalf of the organization and get reimbursed for that. A product like Fyle enables employees to capture receipts in different ways and send them to the accounting department. Since more than 90% of business expenses in the US are done via credit card, we have a very strong credit card reconciliation capability that allows the Finance team to reconcile receipts with the card statement. This can then be appropriately processed by Finance for reimbursement to the relevant parties.

Moderator: Got it. From a scale perspective, could you also give an overview of how big the current Fyle marketing team and organization is?

Siddharth: We have around 80 people. The marketing team is about 11 of us, including product marketing, content marketing, demand generation, and marketing design.

Moderator: Okay. From a go-to-market perspective, is Fyle more inbound or outbound? What would a typical ACV range be like, and what about your sales cycle length?

Siddharth:  We are primarily inbound right now, and just starting to run experiments around outbound and ABM (account-based marketing). Since it’s a very horizontal space, our ACV ranges from $3K - $65K, and our sales cycles last anywhere between 2 - 8 weeks.

The way we differentiate from the market that we’re not mobile or desktop specific like older generation expense management software. We’re in all the apps & devices that our customers use to Fyle expenses.

Moderator: So from an onboarding perspective, is it high-touch with integrations into your accounting system? Or does it work on a standalone basis?

Siddharth:  So there was a time when standalone onboarding would not work. We did not see anyone convert to customers this way. But recently that has changed. We built out really good integrations with the Quickbooks Online, Netsuite, Sage Intacct and Xero, and also made a ton of improvements to the UI and the UX of the product.

So recently, we actually saw people coming in and converting without any customer success touch, but for the majority of our customers, we have a high touch model driven by the customer success team. 

Any good expense management system will have to map to the customer’s policies. So that takes a bit of time and it could also depend on each separate project undertaken.

Moderator: So when you talk about inbound, what would be the split between paid advertising/organic and content leads?

Siddharth:  Close to 50/50.

Moderator: And for paid ads - is it mostly the high intent ones? Adwords, search keywords, or even LinkedIn and Facebook also?

Siddharth:  Mostly the high intent ones. Adwords, G2, and Capterra.

Moderator: And you mentioned ABM? Would that be more email led? 

Siddharth:  That’s being planned out. But it will be email and Linkedin-led because we've tried ads in the past - there were some discrepancies over there. For example with Rollworks, you have to be really smart about it, and it takes a bit of time. I've had conversations with other marketers who've got this to work. And from what I understand, they never used the targeting provided by the vendor. They almost always uploaded their own list, either cold or warm, and targeted that.

Moderator: Between Linkedin and email, from your experience, have you seen one of them working out much better for targeted outbound?

Siddharth:  The cost of email is much lesser, so the ROI on email will be better. But you can do a lot more on LinkedIn, and can actually nurture them back. This is because there are multiple opportunities to get in front of someone. So I can't say which one is better, but the ROI on email is better.

Moderator: Right. Could you compare and contrast with other SaaS companies on organization structures? Why is the organization set up this way as compared to other structures? And how would you measure the North star metrics or goals?

Siddharth: I’ll start off with product marketing. Whenever a campaign has not worked out in the past, for me at least, it almost always comes down to the messaging. It’s about not being able to connect with the audience. I see that mistake happening over and over again because it's very easy to jump into LinkedIn or another platform, create an ad and run it. But it takes time to really talk to people, prospects, and go through sales calls. Maybe we'll get a cold user testing session or a JTBD session - and these things are generally very difficult to do. But they provide the most value.

In some niches you have to talk about the problem first, and then the solution. But for example, in expense management, the problem is very well known. Expense management has been around for probably 50 years and expense management software has been around for 20 years, at least. So, you can show them the solution and talk about it. This is not true in other niches - like in A/B testing. In Wingify, we had to talk about what A/B testing could do for you first, before presenting a solution.

North star metric for product marketing usually comes down to website conversion rate because that is the outcome. If the messaging is crisp, your website conversion rate from landing pages should go up. That is a trend line that we try to look at. But in SaaS, your traffic is less. There's not a lot of traffic like in B2C. So I'd also look at a lot of qualitative feedback. For me, if I can get three user testing sessions done on our landing pages, and those people say that it spoke to them, I have a lot of confidence in our messaging. 

On content marketing, there’s an SEO and Inbound plan. In SEO, I’m looking at keyword rankings, where our landing pages are ranking, organic traffic, and finally, organic pipeline. Demand gen is all about pipeline. In our case, it's a marketing pipeline, which is one step before the sales pipeline.

Moderator: For this marketing pipeline, what is a typical conversion rate from marketing to sales? Is there any issue in terms of marketing teams calling it a pipeline, but no prior buyer intent expression being recorded? Or do you already do that qualification before you put it under a ‘Marketing Pipeline’?

Siddharth: We actually shifted from leads to pipeline recently. We know that to hit our revenue goal, we reverse-engineer from revenue, going back up to sales and then marketing pipeline. Since the end result is a revenue, it works.

Moderator: So the demand generation team works on the marketing pipeline. And there is a drop between the marketing and sales pipeline.

Siddharth: Yes. And especially since we’re not an established brand in the US yet, all buyers are aware that every marketing team makes their products seem like they’re the perfect solution. So smart buyers start out at a percentage of the expected buy order first to try it out at the beginning. But the good thing is we see a negative churn i.e., we see a lot of active users being added to our customers' plans through the year.

Moderator: Understood. So, anything beyond demand generation, content, and product marketing teams?

Siddharth: Yes. Marketing design works closely with product marketing. And generally speaking, their efforts are very synchronous. Then, there's ops engineering. So marketing design and ops engineering obviously don't have any numbers right now, but we're still figuring that out.

Moderator: Got it. And just to compare and contrast other ways that marketing teams are set up in SaaS, are there any specific trade-offs that you made to say that this is possibly the best way to organize the team? In your experience, what would those learnings be?

Siddharth: In some organizations when ABM becomes very important, especially higher up the value chain and the higher ACVs you move to, there it becomes more of a “revenue team”. This is essentially demand gen and sales and sometimes demand gen plus SDR. But for an average SaaS company, which is inbound driven and has a lower ACV of say $3K - $60K, this structure seems to work really well.

Moderator: Understood. And today, how does the demand generation team work? Could you just walk me through that flow?

Siddharth: So I'm a firm believer that any marketing qualified lead should themselves indicate that they want to know more about the product and there should be no ambiguity over this. That is why, for us, a marketing qualified lead is only someone who fills in the ‘Request a demo’ form. For the blog or for a newsletter form, they're a lead in the system, but they're not yet something that we can call marketing qualified and pass on to the sales team. If you see our ‘Request a demo’ form, we ask about the number of employees - as that is the most important thing for us.

Moderator: So, shifting gears. When I speak to other SaaS leaders “marketing source pipeline” and “marketing influenced pipeline” seems to be a very common thing now. How does that apply to Fyle? There seem to be a lot of different interpretations of this, so we'd love to hear your thoughts.

Siddharth: We have a very active content team and plan. We actively track all our rankings for all the business keywords by commercial intent. We put in a lot of effort to rank them higher, to get backlinks, and to optimize the content. So if someone were to say organic is not marketing sourced, I would disagree with them. If the vast majority of our leads were coming in from brand keywords, and there was a separate brand team, I would say yes- it's probably from the brand team, not really from us. But in our case, anything that does not come up from outbound, sales outreach, or the partnership is a marketing sourced lead.

Moderator: Right. So are there any operational issues in measuring this? Here’s an anecdotal example. A person from a company was visiting the website, through some paid ad. They maybe went through a bunch of blogs and were keen to start, but at the same time, another person in that company reached out to the salesperson. Have there been cases where it's hard to measure this or agree on the influence of marketing?

Siddharth:  That's actually a great question. And I haven't faced this problem, at least in my career, because Wingify was very inbound driven when I was there. Fyle is also very inbound driven. I suspect as we move towards higher ACVs, say $100K and above, this problem will come in more often. To be honest, I think it will be situation-dependent, but I can appreciate why a marketing leader at this stage will want to talk about the Marketing influenced pipeline. If they've made even a small dent on a $50k to $100K deal, they’d want some attribution over there.

Moderator: Right. The context of products like Outreach coming up for sales folks to automate an email sequence, as well as marketing teams doing similar things, is where this becomes even more problematic too. So, there are some conversations where sales leaders talk about contacts already being in the pipeline stage, from an email that was sent out. Or the person perhaps downloaded a white paper and that influenced them to convert it. In these cases, how do you think people should agree on a framework to operate in this?

Siddharth: My personal experience is that this is more of a situation where leaders were fighting it out against each other. See in the buying journey, there are multiple influences. A person telling you that you’ve got something interesting is obviously huge. Then you'll do some research to see what this prospect does? Then the salesperson will send you a whitepaper, something of the sort.

So, the buying journey has elements of both sales activity and marketing influence. I've noticed that when the sales and marketing leaders are not aligned and not working together, you have a lot of operational issues. But when the two leaders are aligned, these issues disappear. 

I think by adding a CFO into the mix of a VP Sales and VP Marketing, things change. The CFO needs to know who's responsible for spending the money, and what ROI they end up providing.

Moderator: So the situation which becomes problematic, at least my understanding, is when you have a CFO who says, ‘Hey, this is not really working out. We have to cut the budget over there’. It just becomes complex in that sense. 

Siddharth: Yeah, fully agree. That happens because the CEO or the board will tell the CFO about optimizing the spending. That person comes looking for where they can optimize. And they'll try to cut out waste. Now the problem with a lot of marketing spend is that you really can't pinpoint everything. There might be a ton of tracking software, but once you enter the ad tech space, you really can't see how they got influenced and what ads they saw. Especially on display advertising. Once you get into display advertising and retargeting, there’s so much going on that you really can't tell how much money you can put over there.

Moderator: Actually, that leads me to the next question. In addition to the whole trackable part of the ecosystem, you do brand advertising as well, right? Or brand marketing with podcasts and other things. How do you think people can measure the impact of that? Is it worth even going down that discussion, or is there at least, even some heuristic way to say ‘Yeah, it’s working fine’?

Siddharth: We once tried this through Google Surveys - we really targeted a city. The survey basically asked a version of, ‘do you know what the product does’. We then put down a lot of targeting conditions based on our Google Analytics demographic data. Around 16% to 20% of them replied positively. We then wanted to track this data and build a trendline, to show that we've grown it into 25%. That was an attempt at a tracking brand. But the other way to track brand is organic searches, Google Trends, and your organic brand traffic going up. Those are attempts at putting a number to what is inherently difficult to measure.

Moderator: Right, understood. What do you think about asking the SDRs to explicitly ask a question saying, ‘Hey, how did you first hear about us’ or something along those lines?

Siddharth: Super useful, but often the SDR tends to forget to ask about it if a trend hasn’t been established. But it is super useful, and we do this at Fyle.

Moderator: From an analytics perspective, and given our product is in the analytics space, how has better analytics helped in terms of improving marketing? Would just love to understand your perspective on whether investing more in analytics is something that could change the game for a marketing team.

Siddharth: Analytics helps big time. You just need to know what to look at and where to find it. So for example, let's say an AdWords campaign is not working. It could be that the ad copy is not good, or it could be the landing page not optimized, or even that when people sign up on the form, there is a delay due to many background JS calls. So generally speaking, with all the tooling available today, there is a lot of data, but you need someone who can really dig deep down and figure out what is not working. That is a huge challenge in analytics. This is why generally speaking, there's a tech-enabled marketer today. The people who're very comfortable with HTML, JavaScript, and maybe even some SQL - those are the people who can do the best marketing analytics.

Moderator: Now, if you reach a place of stability - say when you can track a lot of metrics - there's still a lot of volatility in terms of new campaigns, new pages, new offers, right? So how do you stay on top of all of that today?

Siddharth: When you reach that stage, with that many campaigns ongoing, you need a person who's dedicated to that. And at least to me, what happens is wherever a major budget spend is happening, that catches my eye first. And the way we look at that is the Cost Per Lead. Whenever this fluctuates, it stands out and gets our attention first.

Moderator: Right, got it. Last question from my side Siddharth. After that, I’ll just open it up to the audience. If you had one thing on top of your mind from an analytics challenge perspective, what would it be? It could be something as simple as taking time to solve specific problems. Perhaps something has been nagging you for months.

Siddharth: I've actually reached a stage where I don't think of analytics as a problem per se. I've just accepted that this is how it's going to be. Only a few days ago, when I was creating a report in HubSpot, we still had to come down to Excel. You pull in data from different sources and put it into Excel to run through your custom formulae. I think that’s just how the world is, and I've accepted that all my real reporting will probably happen on Excel.

Moderator: So today, when you're using HubSpot, are you able to look at people from a certain industry, or people who have, let's say 11 - 25 employees against one with 300 - 500 employees, to identify how they behave differently? In terms of the different campaigns they respond to? Do they go to certain content as compared to the other group? Is that something you've looked at, and is it easy to do?

Siddharth: It's not easy to do - HubSpot does not have GA capabilities and GA does not have Hubspot connecting - the two of them are disparate systems. Our product marketing team does a weekly analysis on this. So maybe over time, they'll be able to give us some insights. But this is one of those things that we want to understand. How do our larger prospects behave differently from the smaller prospects? We know that the larger prospects tend to check out the website, product pages, and solution pages more deeply. The smaller guys convert to MQLs quicker, and also close quicker. In this situation, I don't know if there's a deeper insight that I'm missing out on.

Audience Question: On a weekly basis, just to get a sense of how things are working - what are the metrics you look at? And how much of a help is it for you to figure out the next course of action?

Siddharth: So whenever we decide on a plan for the upcoming quarter, we also decide on the numbers we're going to look at and track. In cases like SEO week on week, we'll look at our keyword rankings or our organic traffic. We’ll then add organic conversions and pipeline - that tracking system gets installed at the beginning of the quarter. And we have an analyst who helps us set all of these up. I usually tell him about what we need to do and get him to set it up for us. We then look at it every Friday in our weekly team meetings.

Moderator: Do you use HubSpot for this, or do you use Google Analytics? Or do you end up monitoring the ads metrics on Google ads and Facebook ads, and then the website metrics on GA and HubSpot for the post-lead metrics?

Siddharth: Exactly the latter - we use different systems to get different data points and put them into an Excel sheet. For example, the content marketing team will show data from SEM Rush, Ahrefs, Search Console, and Google analytics, Demand Gen will show data from Adwords or Linkedin or Reply.io, and Product Marketing will show conversion rate and behavioral data from GA and Hotjar.

Audience Question: Just a follow-up question - how often is there a need to dig deeper into data to explain something? For example, ‘the leads don't look right’ or ‘there’s a sudden increase in leads’ or something of the sort?

Siddharth: There’s no definitive answer here - but it is triggered when we make a change in something. Say, for example, the product marketing team has pushed live some new pages. They want to know how this is impacting visitor behavior, but after two weeks or after a month when a new baseline has been established, then we don't look at it. We know how those have impacted the outcome. If there is a new Linkedin campaign going on, we'll track it every day. But let's say an AdWords brand campaign is also starting. For AdWords, it’s more about setting it up, optimizing it, and then reaching an acceptable cost per lead. After that, you just keep looking at the CPLs every week but don’t really dig deep into it everyday.

Audience Question: I wanted to understand a bit more about what you meant when you said that marketing at Fyle is different from marketing at other places. You mentioned that the problem description is straightforward and you have to stress more on the solution. Could you please elaborate on that? And how would the metrics you track for these cases vary?

Siddharth: So, take AB testing for example. When I was in Wingify, at that point in time, you would have to explain that conversion rates are low and that here's what could be done to change it. You'd have to show the problem first. Today, most of the marketing world knows about this - you don't have to explain AB testing or the value of AB testing to them. They just talk to you about the solution they provide.

In Fyle and the accounting world, expenses have been maybe more than a hundred years old. Employees have been making expenses on behalf of the organizations. And this problem is very old. So it’s not a question of leading with the problem, but instead leading with the solution - your product.

In a situation where people are figuring out the problem and the solution, the sessions per user, that is how many times a person comes to your website, is very high. They'll come to your website on four or five different occasions to really figure things out. They might be pulling out case studies to show it to their management and things like that. In cases where the problem is well-known and they're looking at the solution, they will mostly come only once to your page. They will go through multiple pages - maybe 7 - 9 of them. So the pages per session will be very high, but the number of sessions per user will be lower.

And generally speaking, for known problems, you get budgets allocated easily. For unknown problems, budgets are not given that easily.

Audience Question: For a company in the $5 to $30 million ARR range, what are your views on which type of attribution to focus on? Say between first to last touch, and multi-touch attribution?

Siddharth: So it depends on a lot of things. For SMB focused companies, first touch attribution is usually more important, because the sales cycle is low and the number of campaigns they run is much higher. Where the sales cycle gets longer, in B2B spaces with higher ACV towards the enterprise level, the more that multi-touch attribution starts to matter. More conversations happen within organizations to proceed with a $150,000 investment, so multi-touch attribution becomes far more important.

Audience Question: My second question is around giving a quality score. Every company would have its own customized model for lead scoring, but if you were to point out a few attributes that will remain constant for all the B2B SaaS companies, that would be interesting.

Siddharth: In my opinion, you can't have a model that would be a good fit for all B2B SaaS companies. I personally wouldn’t attempt that project because I know it would not be too successful.

Moderator: One last question from my side. One of the sales leaders that I spoke to, said that all of the costs that the sales team incurs with setting calls could be loaded up the marketing CAC with all of the subsequent costs that come in. Have you thought of it in that sense, that there is a huge expense on the sales side, which is both the opportunity cost of not working on the best ones as well as getting on calls and going through the process?

Siddharth: I think that is generally pulled into the sales and marketing percentage that is spent from a company's revenue. That’s where that basically gets counted in. 

Moderator: Thanks a lot! This was very useful to our audience - it was great to have you here

Siddharth: I’m happy to help. Thanks for having me here.