Chris Strom:
Welcome back to the RevOps Hero podcast. On today's episode, I have on Emma Cimolini as our speaker. Emma is the former head of marketing for multiple B2B SaaS startups. She's currently working as a fractional growth marketing leader for an incubator and VC firm called Highbar Enterprises. She's also worked for publicly traded companies like Microsoft Canada and Informa. She's helped grow ARR from six figures up to over 4 million. She's built and launched multiple growth teams and learned how to build great and deep relationships with B2B sales teams as well. And she's been using HubSpot as her CRM of record for doing all of these things over the past six years. We're having her on here today to talk through how she does all that. So Emma, thanks for joining us and welcome to the show here.
Emma Cimolini:
Thank you very much. It's a pleasure to be here.
Chris Strom:
We'll be talking about using HubSpot for PLG companies, so product-led growth software company. As we're talking through that, the first thing we'll start with is talking through the systems architecture of everything. What are the different systems involved and how do you like to go about connecting them all?
Emma Cimolini:
Yeah. So systems architecture for marketers out there, it's really just your marketing tech stack, right? So you want to think about, you've got your billing tools like Stripe is a really common one, your CRM. You probably, for a product-led growth, you want your in-app analytics. You're probably going to have website analytics. Maybe that's GA4. Maybe that's something like Mixpanel that can do both. You'll find what makes sense for the type of data that you want to pull out, but you have to map out what are your different systems and how you want to build or integrate these systems. So that's kind of what you say with systems architecture is what data is pulling from which tool. And a great way to start, it can be quite overwhelming, is what questions do you want to answer? And that will help guide what kind of data you want to connect with what platform.
Chris Strom:
Okay. A lot of the components are things like your app, database itself, app analytics tools, website analytics, the CRM system, the billing system.
Emma Cimolini:
Customer success as well. So things like Intercom probably want to be connected in as well.
Chris Strom:
Okay, so those are some of the common components. And then what so the connections between those systems look like? What sort of data and objects are you connecting between the systems and what are you using for the integration tools between them all?
Emma Cimolini:
So when it comes to integration tools, this is where I partner quite heavily with the engineering teams because you really want some support. They might set up segment, maybe you're using something like Google Tag Manager which a marketer can easily do. But oftentimes you need the support of the engineering team to set this up. In terms of which data that you're pulling. It really comes into a question of, in the case of HubSpot and the CRM, how much you want to rely on HubSpot as your source of truth. And that will guide how much data you put in there. In my experience or how I've done it in the past is when you're working at an early stage startup, you're not ready for say a dedicated data warehouse or a data lake or something like that. You don't have the resources to invest in that.
So a good way or a good alternative in the early days is to use HubSpot or a CRM of any kind to be your data warehouse. So that becomes your source of truth. And what I've done is just taken all of these different systems that have touch points with the customers from billing to their support tickets to which marketing channel they found the website from. Everything gets piped into HubSpot. Everything kind of flows in together. And then the crux of all of that is just how do you organize your data within your CRM, like HubSpot? So that's your data architecture because HubSpot, just like any CRM, it's a relational database. So you have different objects. So your object could be a company, it can be a deal, it can be a contact. And you want to understand how each of the objects connect to each other.
And you want to create a system because HubSpot is quite flexible. So you can align it to how your business flows or your business model fairly well, but you need to be conscientious about what that is. If you let everyone in your teams have access to it... It helps have one person own HubSpot. Everyone can access it, that's fine. But have one person define the definitions, that's very helpful. If you are a B2C customer, it's likely you will use contacts as your primary object. But if you are a B2B, it's probably going to be your company as your primary object, and then you have your contacts within it are associated.
And then from there, if you're using HubSpot for sales, then HubSpot's a really good sales tool. Not the best, but it's very solid. You have your deals associated with each of these and these are three distinct objects that will become the pillars of how you manipulate your data and look at your data. For example, this contact found the website through this marketing channel, but this other contact found us through this marketing channel. What does that mean for the company that they're associated with? So it really depends on which one is your primary object to really understand the holistic marketing journey.
Chris Strom:
And then you're talking about you like to use tools like segments a lot of times for the integration layer between the tool.
Emma Cimolini:
Segmenting absolutely can be that integration layer. It really just depends on what the tech stack is with the product because when you're doing product led marketing then is very much integrated into the product ecosystem. So it might be a bit of a lazy [inaudible 00:06:39], but I simply go to my engineering team and say, "What makes sense for how you guys are functioning?" And then we fit HubSpot within that. So if they're using Segment, absolutely. If they're using just native integrations that they built themselves, it's going to be look like something like that that might create from technical debt further down the road. But sometimes you've got to balance speed with perfect execution.
Chris Strom:
Yeah, okay. If they're already using Segment, maybe for some other uses, you think it oftentimes makes sense just to go with that versus if they already are using API connections to other tools and that's what they're most comfortable with, then just stick with that.
Emma Cimolini:
Yeah. That's what I would lean into. At some point, when a company gets to sufficient size, likely you're going to bring on board like a data engineer where it's their role to build the pipeline because it can be complex. There's a lot of work involved in it once you have more and more systems and more and more data. In the early days, you can be a little scrappy.
Chris Strom:
Yeah. Just start with standard CRM object like companies and contacts and stick with that until... Only go up to a more complex system when you have the need and bandwidth for it.
Emma Cimolini:
Yeah. And there's a lot of the tools that have native just easy integrations that you as a marketer can set up yourself. If I'm remembering correctly, I haven't used Intercom for a few years, at least when I was there, you can turn on a HubSpot integration yourself with a marketer without any engineering support. And then you can control which objects are getting pushed so you can choose how that data is being pushed through. Again, you can put everything in there. I kind of like throwing everything in there. You also can do two-way syncs. For example, if your customer success team is working out of Intercom, there's going to be details about say which plan or pricing that they're involved in. You could build an integration directly into Intercom and HubSpot or you can use HubSpot as your source. So your billing gets put into HubSpot and HubSpot is the platform that pushes it to Intercom.
So you want to think about, you use it as your central hub. So everything goes in and from within that, that almost is your version of Segment in a way where it gets pushed back out to the other platforms. Again, early days. There's much more [inaudible 00:09:14], but it can be... This is how I've done it when it's been a very small team before where you're fighting for engineering support. Sometimes it is an easier way for marketers to get things rolling.
Chris Strom:
Yeah. To get started, just use a CRM like HubSpot as the central data source and kind of the traffic controller, and then later on look at more complex setups involving more platforms and maybe Segment or other data warehouse connectors as well.
Emma Cimolini:
Exactly.
Chris Strom:
And that leads to another thing that we were talking about was deciding between using HubSpot as your source of truth versus using data warehouse is a source of truth. And how would you go about making that decision?
Emma Cimolini:
I think it comes back to just how many resources you have on your team in terms of who can manage... If you don't have a data engineer and your engineering team is really focused on, say, launching that next new feature, they're not always coming back to data integrations. It's not their highest priority at the moment. So in that case, I think it makes sense to use something like HubSpot. Also, because HubSpot, a few years ago at least they switched over to billing you based on your marketing contact, not just the entirety of your contact. So it gives you a lot more control over how much you are being charged, and you're not charged for all your historical data. So unless you're actively sending automated email to a contact, that contact, so long as you mark them as a non-marketing contact, you're not being billed for it.
So it gives you the opportunity to build out your database and you can collect as much data as you want. And you can then analyze that data, you can pull spreadsheets, you can build reports. You have access to it all. It's just you're only being billed at least for the marketing add-on. That's what you're being billed for when you're actively marketing to them. So I think that's an added value to using something like HubSpot, it gives it that extra reach, because one of the unique pieces of product-led companies is that you will have such a high volume of contact and a high volume of companies in your database because everyone's signing up for say a free trial or something like that. It's significantly higher volume than if you were sales-led and you're working on larger deals. So those are usually smaller accounts, things like that. And it's a little... Unfair I don't think is the right word.
I think that's why HubSpot probably set up this concept of the marketing contact, make it a bit more even for large volume companies like PLG companies versus your sales-led because yeah, you need to be very mindful of those usage costs when you're dealing with high volumes like PLG.
Chris Strom:
Yeah. Yeah. So even if you're importing every single free user into your HubSpot portal, you just maybe choose a small fraction of them that really meets your criteria for being really ideal for a prospect and maybe just marking that small percentage of them as marketing contact.
Emma Cimolini:
Exactly. The way we set it up in the past was we actually had our transactional email with a different tool just because we wanted to have just such crisp, clean control over our deliverability. And it was also better integrated with some of our other systems inside the platform itself rather than HubSpot. But that meant that the only contacts that we were really saying were marketing contacts were anyone getting our onboarding email, for example. If they don't take a certain action by the end of the trial, we keep them for a period of time. And if they're still not taking action, then we put them into a cold lift and we use a workflow to automatically... Not to moat them, that's the wrong word, to change their category to non-marketing contact. And so that we always keep just a really tight definition of based on usage behavior, what that looks like.
Chris Strom:
Yeah. Once they become just too cold, you just mark them as non-marketing and save the budget there.
Emma Cimolini:
Exactly. And you can always go back and try to reactivate them, rewarm them. But you're only using that in short spurts rather than just paying [inaudible 00:13:29].
Chris Strom:
Yeah, you can always remark them as a marketing contact later if needed.
Emma Cimolini:
Exactly.
Chris Strom:
Yeah. Okay. So one good rule of thumb is if you don't have a data engineer, don't use a data warehouse, or you probably shouldn't use a data warehouse and just use HubSpot. What would be some cases, what would be some advantages and some good cases for using a data warehouse as the source of truth?
Emma Cimolini:
You're probably going to have more nuance. You'll be able to have more control over your data. So I mean there's limitations with HubSpot in general. So it's possible that you can cause the system to slow down because you're going to hit limits with your API calls. It can be very expensive in general. So granted, if you're a small company, probably not going to be doing data warehouse anyways. But if you need to have really specific data types that go beyond just your standard company, deal, contact, you need more control over what kind of data is being recorded and in what [inaudible 00:14:38]. Then a data warehouse is definitely going to be a company you want to look into. So for example, I'm working on a project right now where we have... It's within the legal space, so there's various types of entities that wouldn't necessarily be well-represented within something like HubSpot because we're building out a database more so than just a...
It is a CRM, but it's hard to go into the details of it, but it's on both sides of things, [inaudible 00:15:06] what I'm trying to say, depending on how complex your model is. One of the limitations I found with HubSpot was when you're working with PLG, you have a subscription model and pricing will often be kind of flexible. So for example, people might upgrade their plan one month, take it down another month. "I want this add-on, I take this add-on off." HubSpot's not really built to track that level of flexibility or dynamic pricing changes.
So if you really need a really tight focus... Everyone needs to be focusing in on their revenue and how that being modeled. If you're using HubSpot as your source of truth in your own warehouse, you'll need a separate tool to analyze your data. If you want to combine those two things between your revenue and all of that, with all the stuff that you have in HubSpot and that's mission critical, then you're going to need to look for a warehouse as well because it'll give you more flexibility. So that's probably one of the bigger use cases I see with what you're trying to get out of your data, to use a warehouse versus HubSpot.
Chris Strom:
I know like CRMs like HubSpot or Salesforce, under the hood they're databases. I've heard it said that data warehouses in general oftentimes are better at databases for tracking changes over time and historical data because databases are primarily made for tracking the current state of the data, not as much the historical state.
Emma Cimolini:
Possibly, yes. Depends on what data's changing. Again, it comes back to your own unique use case, and it depends on what you're trying to track. Like timestamps, for example, can be tricky. That was definitely something that I ran into as a problem when I was at one of my other company. For example, trial signups, we would have people sign up for multiple trial signups. We had trial users. They would sign up for one, then another. And I could never tell based on just how we had structured the data if someone had signed up for five trials, one, two, a hundred. I only knew the most recent time they signed up for a trial and when the contact was created. That was the only visibility I had, and I didn't have that granularity. And that was just a limitation of HubSpot.
There might be a way to solve it that I didn't find, but you might find yourself running up against these types of limitations. So I think it's really great to start with HubSpot because you start to learn what your limitations are and where your gaps are, and then you start to invest in something like a warehouse.
Chris Strom:
Yeah, as you have the capacity and budget to start bringing on data engineers to start doing some of this more complex structuring and pipelines. Right?
Emma Cimolini:
Yeah. I also think when you're in the early days, you can get by with being scrappy and really getting into the weeds of your data. Don't be afraid of a spreadsheet. Sometimes it really is all you need to go from A to B. It's not going to get you from A to Z, but it will get you to that next step.
Chris Strom:
Cool. Well, we've been talking about engineering support needed to build a data warehouse. And of course when you're at a company and you're getting to the point where you feel like it's time that you need that engineering support, how do you go about making the case to get the budget for the headcount or the consultants or other resources to start implementing that?
Emma Cimolini:
I think in the contents of a product led company, it's a bit of a unique case where you're not siloed because it's product led and usually your leadership team are technical. Not always the case in startups, but often they're technical or that's their biggest focus, especially in the early days. And so if you're talking about, "Okay, we want to lead with product, we need our data to talk to each other," you want to bring on board your product team and make sure your product team understands the list and the benefit of leading with product and the amount that's involved in that in terms of tracking not just an MQL, but a PQL, getting your in-app analytics and tying those to revenue as well as what's happening in your marketing activity.
If you can align around that core objective of using that product to grow the company, you talk to your product team, talk to your sales teams, you talk to your leadership, I think if you're aligning yourself with everyone else around that goal, it should naturally come forward that okay, now's the time because this will be a shared resource, not just for marketing, it'll be for everyone.
Chris Strom:
As you're reporting, everyone will gradually start to see the need. Everyone will start to run into more questions that are more difficult to answer as things get more complex.
Emma Cimolini:
I think if leadership has done their job in sense of, "Okay, this is our go-to-market motion, this is our strategy, we're going product led..." Even if it's not product led, maybe you've chosen a hybrid motion, for example, where there's a bit of sales and there's product. But regardless, leadership has said, "This is how we're going to market," you need to be able to answer certain questions and marketing is going to be involved in the product and the onboarding journey and product's going to be involved in helping marketing streamline that friction. So you guys are going to have a lot of shared challenges. And so I think it's a lot easier to make the case bringing on this headcount or this team member when multiple team are sealing the bottleneck, not having them.
Chris Strom:
And also I think that's a good point that you made that a lot of software startup companies are founded by engineering people who built the product. And so because of that, they oftentimes they'll be pretty naturally inclined to hire and it won't be a lot of arm pulling. You might be more often in other companies.
Emma Cimolini:
The best way though I can say sometimes in that case they might lean a little bit too far about. "We have all the analytics in the product, but what else do we need?" You're like, "Other pieces to it." So if you can tie in your marketing efforts and what you're trying to do to revenue and frame it in the context of... Honestly, that's what your CEO, and that's what your product manager is really looking at is like, "How do we grow the user base? How do we grow our user base and drive revenue?" And as a marketer, if you can align your activities to those questions, it also helps to have a much more cohesive conversation.
Chris Strom:
Yeah, yeah. That's a good point that sometimes they might be more inclined to overbuild on the product analytics side and sometimes it might be more of a task to keep everyone focused on the revenue picture more than just product analytics or product signals for the sake of product signals.
Emma Cimolini:
Yeah, exactly.
Chris Strom:
And you had mentioned just now on talking about product signals, you had mentioned the PQLs and things like that. So let's talk about life cycle stages in the funnel for product-led growth companies and how they are similar or different to funnel stages or life cycle stages at sales-led companies.
Emma Cimolini:
This is where your data architecture or how you want to link your different objects become tricky. So a life cycle stage for anyone who's [inaudible 00:22:48] hasn't heard this term before, you go from a subscriber to a lead to a marketing qualified lead with their HubSpot definition to an sales qualified lead, opportunity, and then customer, and then they have another category. The trick inside HubSpot is leads all go sequentially. So they can go from a subscriber down to a sales qualified lead, but the timestamps don't really work if you try to go backwards. It really should flow like a waterfall. In a product-led company, you probably want to change it up a little bit so you can change your definition from instead of a marketing qualified lead. I would recommend using product qualified lead. The marketing qualified leads are often going to be things like, "They signed up for a free trial and they sent a certain criteria for our demographic. They're in the right vertical and they're the right size and they signed up for a trial. Sales, try to talk to these guys."
But when everyone can just come in and check out the product and see what's going on, you get a lot of tire kickers. So they may not be the best lead for sales to chase. This is particularly true if you're using a hybrid motion. Or if you are using marketing as your main driver in trying to use your product, you want incentivize as much usage as possible to show that the product is doing what it needs to do. So at the end of the day, a PQL is your industry qualifications, just like a marketing ed lead. They signed up for the trial, they fit a certain vertical or however you want to define it, but they also are actively using the product. So maybe they've reached their aha moment or they're using certain features to a certain extent based on how they're using the product.
They're showing high intent. So either A, that's a signal for your sales team to reach out to them, follow up with them, talk to them, or it's also a very good indicator that product is working the way it should. It's onboarding them, it's reducing the friction, it's pulling them through the funnel in the way that you're expecting. And then from there, I would just kind of skip SQL unless you have a sales team. If you're product led, just skip it. PQL is sufficient. If you do have a sales team though, having SQL is simply like a sales person has talked to them and [inaudible 00:25:13] qualified them, your standards of sales qualifications. They have the budget, they have the timeline, that kind of thing.
And then if that's the case, they go into opportunity and closed one and so on. If you are product led, I would skip opportunity and SQL completely and just go from a PQL to a customer. But then in any kind of subscription company, you also want to factor in churn. In addition to life cycle stage, you also have something called your lead status. That's the [inaudible 00:25:40].
Chris Strom:
Oh, lead status.
Emma Cimolini:
If someone churned, I actually keep them in the customer life cycle stage of a customer and then just change the status to churned. So learning how you want to structure your data, everyone will do it slightly differently. Whatever makes sense again to what your needs are. It really comes back to how you want to use the data to answer your question. That'll define how you structure it, but that's how I've done it in the path. And then the other thing to be very mindful of is that you really want to be very clear over specific trigger events inside HubSpot.
So a trial signup, this is where you really need your engineering support. So inside the product, in their backend, however they've chosen to set it up, a trial signup happens in the backend. You make sure that it gets pushed into HubSpot as a unique property. Then you can tie your MQL or your PQL status to this product event. Same with a churn event or a customer, an upgrade event. So you want to make sure in these very key linchpin pieces, you want to make sure that what's happening, the product analytics and HubSpot are a one-to-one. It doesn't have to be for everything. But for these really specific pizzas that are so important to that overall flow, you want to make sure that's set up correctly.
Chris Strom:
Make sure you're logging the key events like signups or cancellations.
Emma Cimolini:
Yeah, exactly. And just so your HubSpot is a mirror reflection of what's happening inside the actual product.
Chris Strom:
So someone could go in, pull up a contact record in HubSpot and look at it and get pretty quickly understanding of their current status.
Emma Cimolini:
Yeah, exactly. They signed up for this trial on this date, then they upgraded to a customer on this date. But then they churned, but then they came back. You want to be able to kind of understand what's going on.
Chris Strom:
Yeah. It's like maybe log those events, timeline events on the contact record.
Emma Cimolini:
Exactly. And you can create a unique property inside HubSpot, and then you can create workflow inside HubSpot so they trigger a certain lifecycle status or lead status based on the trigger event.
Chris Strom:
And then what do you oftentimes do with those? When you're using the lead status to track churn people, do you sometimes put them in an email workflow to try and re-engage them or maybe assign them to a sales rep to try to connect with them or something else?
Emma Cimolini:
Yeah. Both very valid options. You can try and put them in an email dress, put them in a list for your sales guys to come back to in six months or whatever it might look like. You can put them into your... maybe you have a certain set of ad campaigns, and so you're pushing that into LinkedIn or as a retargeting audience inside Google Ads or however you want to work with that, right? There's so many ways that you can use those lists. You can give it to customer success simply to say, "Hey, go and find out why they churned if they didn't do a customer survey when they entered it."
Chris Strom:
So those are all possible things you could do. And depending on your use case, you would maybe choose one or a couple over other ones.
Emma Cimolini:
Exactly.
Chris Strom:
And I know in my case, I oftentimes end up never using or even just removing entirely the subscriber stage and also the evangelist stage myself. I usually end up just using other metrics. If I want to track those, I usually just end up tracking other things instead as well. Do you like to use those two stages yourself, or do you oftentimes get them?
Emma Cimolini:
It depends. I guess it depends on your. So if we do genuinely have a newsletter, right? Blog subscriber list, something like that, I genuinely use the subscriber status of anyone who has once received content from us but has not had any kind of meaningful buying intent interaction with us. Then I leave them as a subscriber. And only once they've shown some kind of intent, whether they book a demo, contact form, something like that, do they become a lead. I might in some cases keep them as a subscriber. If it's like an outside lead list, maybe you found some lead through Apollo or something like that and they've been imported, I would keep them as a subscriber rather than a lead because they haven't engaged with us. If they replied to us, I might call them a lead. So it really comes back to just how you want to structure your data, how your brain works or how your business works. Just removing it entirely can also be very valid as well.
Chris Strom:
Yeah, depending on how you want to set the criteria for the different life cycle stages.
Emma Cimolini:
Exactly. And evangelists could be just like anyone who's given us a testimonial or a case study are an evangelist and [inaudible 00:30:54] the manual. It's just easier to find them.
Chris Strom:
Yeah, based on maybe using the NPS score.
Emma Cimolini:
Yeah, exactly. I've done that in the past too.
Chris Strom:
Cool. Well, let's see. For the final question here is, are there overall any things you would recommend being mindful of or watching out for in terms of the strengths or limitation of using HubSpot in a product-led growth company?
Emma Cimolini:
Yeah, just a few of the things I think we've chatted a little bit about was just the price, especially if you're a smaller company. HubSpot's great, but it does get expensive very quickly, so be very mindful of your marketing contacts versus your non-marketing contact. If you have large volumes of data, you might see the system slow down just with there's limits on your API calls and things like that. So another sign that if you start running into that, you might want to consider doing your own data warehouse. If you're using it for sale, I think it's a solid sales CRM. But it's not necessarily the best. For example, it doesn't have auto dialer and things like that. So there are some limitations. HubSpot, it gives you all the tools, but its best tools is your CRM database and the marketing automation. Sales is decent, customer success, customer support has some work to be done.
The reporting functionality when you do custom reports is excellent. I do actually really enjoy that function of HubSpot. So it just depends on how you want to structure and what that price point makes sense for you. You're still though going to have to separately build a growth funnel in a spreadsheet. These are our channel, this is the percentage of the traffic, these are our leads. You're still going to have to separately track all of these things, which are important. And I guess my biggest caution would be just be very intentional about how you structure your data because it does get messy very quickly. It's good to have one person sort of own the data hygiene and definition. It doesn't have to be a dedicated resource, but just someone who knows how all the systems work. It's very helpful. And start small and grow from there. If you've got absolutely nothing and you're starting from scratch, start small. It will grow over time. And yeah, I think that's my pieces of advice for people who are just kind of getting started.
Chris Strom:
Cool. I think there's some great content we covered here and great things to think about here. So Emma, thanks so much for taking the time to join us on the show here and hear all of these things you've learned from your own experience.
Emma Cimolini:
Yeah, it's been a pleasure. Thank you so much for having me.
