How to Build Your Marketing Stack

The panel responded to questions from the audience. The first question was… What's the essential marketing stack every SaaS should have; regardless of size?

Mr. Cabane said, “For me it really starts by having some kind of data collection; being able to understand who's on your website, who's on your app, what are they doing? And so, I spent the first two months building a deep-tracking plan to understand everything that we're doing right.

And then I need to be able to do some analysis. So, I plugged in Amplitude, and then they need to reach out to those people, based on the analysis. What I do is break down my audience in different buckets - based on the behaviors, based on the demographics - and then, I try to customize their experience. How do I customize my stack? Emails, in-app messaging, onboarding. So, it could be some app cues. It can be some Drifts. It can be some customer.io. I can touch multiple channels, but that's basically my stack. So, data collection, data analysis, and data action is how I break it down.

Pauline, [Do] you have a perspective on this question?”

Ms. Fumeron said, “Data for me is essential. If you are not able to know who is coming to your website - what's working, what's not working, and what's going on with your product - it is going to be very difficult to make choices with marketing.

I think it's essential that you are always able to get feedback from your customers and also to engage with them with the right message at the right time.”

Mr. Cabane said, “Alice, I think you have an interesting perspective because you're at the very early stage, and you need to engage for sure. You're doing one-to-one marketing. Do you collect data? Do you do any analysis at your stage?”

Ms. Default said, “Yes, we do. We have data everywhere. So, we have Tableau, we use Mode; which is kind of like Amplitude…”

Surprised, Mr. Cabane asked, “You have Tableau in a pre-product market fit company?”

Ms. Default responded, by saying, “Yes. Our Ops Director - who's in charge of data - loves Tableau.”

Mr. Cabane asked, “You don't have many users or customers, so what do you utilize in Tableau? Is it your own data, or is it market data?”

Ms. Default said, “It's our own data. We have it connected to our products. We have people using our apps and right now we're mostly just tracking feature usage; which features they are using in the app? Which ones should we push for? Which ones should we kill? We're killing 50% of the app this week because we realized that it wasn't necessary. So, we use it just to help the product mostly.

And then Mode is more [for] trends. Like, how do people use specific features? Do they come back to it? How's retention? Things like that.”

Mr. Cabane said, “Let's go to the second question. How to measure marketing activities in B2B sales. And any tips when marketing enterprise?”

Mr. Cabane addressed the question by saying, “My take on it, at my stage, was to be able to understand the number of interactions with the [enterprise] accounts we're talking to over time. So, how many times have we touched the account over a period of time? And, we do regressions to understand how touches impact our conversion rates. What we saw, which is fairly obvious, is that the more touchpoints interactions we have, the higher the likelihood to close.

It's called conversational marketing. When I talk about a touch, it's not sending an email - because that has zero value - it is... Did the end user receive value from my touch? Which means basically that they responded to my email. Did they call us; were we able to talk to them? Did they come to a booth at an event, or did we have a chat with them?

So, did we have a conversation? What is the length of the conversation and how frequent is that conversation. The name of the model is RFM. It's a Recency and Frequency Model. So, if we have five conversations in the past 30 days, that's good, right? It's really good. And so, that's how we were measuring marketing effectiveness: the ability to generate conversations and
responses.

Alice, I think you have some experience in that from your days at Sunrise?”

Ms. Default said, “Sunrise was B2C, but Front was B2B sales. We had shorter cycles. On our side, we did a lot of content, like newsletters... You stay very present with your lead until they make the final decision. Um, so contacting them on a frequent basis, going back to them, sharing new updates, new features.”

Mr. Cabane added, “The goal of the email, at the enterprise level, is to get the response. So the email comes from the sales leader, the sales rep, or the CEO; depending on your company size. The responses are automatically analyzed by MonkeyLearn; which is an AI tool. So, I can do some sentiment analysis. Is the response positive, negative, neutral?

If it's positive, it goes back to the sales inbox. If it's neutral or negative, it goes to the trash. And the goal of that is to understand... are we creating a high volume of positive responses? And that is super scalable.

Let's take the next [question]. Do we have any golden rules for hiring someone versus buying a platform? Alice, I want to start with you because you asked me exactly that question like 20 minutes ago.”

Ms. Default responded, “I don't have the answer, but I have to rephrase the question... We are a super small team. We're mostly focused on product right now. I know my next role has to be some sort of marketing role. We've hired on the product and engineering team a lot. The question is... I don't know which role to hire for because I'm not sure which product I'm selling. I'm not sure who I'm selling it to.

We're mostly experimenting right now. We're doing events, we're doing a tiny bit of content, we're doing community building partnerships. And I don't know if I should just give this to an agency? Should I hire someone internally? How do I make the decisions?”

Mr. Cabane said, “My take on it is ...I never hire someone to rebuild a specific product that already exists. I hire people who are able to customize complex tools. HubSpot is a good example of that use. Zapier. Having a human that competes with an entire company - which is dedicated to building a product - it's a losing battle. There's no way I can hire someone to compete with HubSpot.

So, I try to hire people who can improve existing products and build relationships with those tools.”

Mr Cabane addressed Ms. Fumeron, “Your [marketing] team has grown from three to eight people in the past six months. Maybe you can tell us a bit about how you decide [between] build versus buy.”

Ms. Fumeron responded by saying, “It's actually very interesting because we are - more and more - doing things internally. But, we are making sure that the marketing team is also being able to make the connections with the different tools and teams. For instance, in our case, I've been able to connect our CRM tool with the customer success teams. We're able to proactively contact students if we need to. Also, it's a matter of seeing that all the value, that we have somewhere, can be shared.

Having people with more and more competencies - experts in a field - but also being able to have a step back and understand how different tools work, and how you can link them, is very helpful. Rather than being good at just one thing… how can you go further, and how do you create value with all the tools and data that you have?”

Mr. Cabane said, “The next question is... Most companies don't invest in the data infrastructure segment, in the beginning. When do you say it's the right time to do so?”

He addressed the question, by saying, “At the beginning it's the right time! Because it's not so much a tooling problem as it is a culture problem, right? It's always easy to hire consultants and say, We're going to put tracking everywhere in the app. But, you haven't really solved the problem. They don't know how to preserve it.

It's impressive that I see them being used from very early stage to... Segments still uses it. Drift still uses it. The downside of Zapier is that when you have like 200 people building zaps, it's really hard to understand what the data flow is.”

Mr. Cabane moved to the next question, saying “What are some ways that marketing and customer success can work together? Pauline, [will you] take that one?”

Ms. Fumeron said, “In our case, the first [step] was talking to each other. At OpenClassrooms we are always in a little bit of a fight. For instance, every time we launch an email marketing campaign, the customer success group comes in and says, Hey, we're going to have tons of tickets, it's going to kill our KPI... But we've been able to slowly understand each other's needs and to find ways to reach each other's objective.

For instance, when we are launching new projects, we are getting the student success team in [to understand our] impact on their work. So we're making sure that we have focus in advance. What were the needs? What can we improve?

For instance, the marketing team developed a system of alerts so that customer success can proactively reach out to customers. And, this was not possible without the help of the CRM. So just building the bridge and talking to each other for me is the key to success.”

Mr. Cabane said, “Let's move to Emil, who's asking a question about OpenClassrooms. How do you work with marketing automation, et cetera?”

Ms. Fumeron said, “We have different people working in their expert field. My job is to create a link. For instance, when we have new acquisition campaigns that are being launched, how do we make sure that the people have the right tracking, receive the right message, [and] the sales team has the right information about the lead. The mission is more about collecting and enriching the data, so that we all have clear information about the leads and the customers that we have.”

Mr. Cabane presented the next question, saying, “In an early stage startup, with no current marketing hires; how would you advise the first dimension hires to spend the first few months?”

He addressed the question, by saying, “If you're talking early stage, pre-product market fit, the job of that marketing hire is finding deep market fits. An example of that is... that person should be able to build 20 unbounce pages, spend $2,000 on AdWords and Facebook, drive traffic to those pages - which have different copy, different appeal, to different audiences, different messages - and understand the KPIs of what's the click through on my ad. What's the conversion rate on my unbounce on my form? And who has actually converted?

Understand the size of that audience and then give back to your CEO information on: This is the size of that audience persona. This is how much it costs to reach each of them. This is the quality of the audience. And, here's the message that appeals best to them.

Because building a product is one thing, but imagine Alice builds a product, then she wants to acquire those users. Isn't it better if she already knows exactly the metrics, the cost of acquisition, and the conversion?

One of the best strategies is often to build a fake payment page - if you have a product that requires payments, right? There's actually no payment happening, but build a fake minimum page and see what is the on-payment-page conversion. Does your message and your fake product resonate well enough that people are willing to pay for it? If the answer is yes, you have a product, just go and build it, right? If the answer is no, then you have a challenge.

And that reduces the cost of spending on products. Here in France - and I'm French as you [can] hear - French, people tend to build way too much product before they go to the markets. WAY too much product. So build less product and go to markets sooner.”