Falkon PQL Workshop
By Elizabeth Gallagher
June 2, 2022
0:00 / 46:44

So just as a little bit of the, what you're going to see here is essentially a summary of all the things that we as practitioners have uh, at our um, at Dropbox at one drive and other companies that have. Had a strong product led growth motion from free to uh, and also us working with, uh, a through uh, that have a strong, free paid, uh, workflow.

So this is sort of a set of best practices that we've identified.

PLG vs traditional model

So high level before we get into the nuts and bolts of you know, what are the key differences between a traditional go-to-market motion versus a product led growth marketing motion? It's not simply, you uh, success. Right. Um, In a PLG model, we also need to customize our outreach based on.

The person we're reaching out to and their usage of our product. For instance, if you're reaching out to someone uh, a marketing, a growth motion.

You have free customers and in a traditional model, you don't, it's not as simple as that. So I wanted to highlight about. Key differences and then we'll deep dive into each of them. So the first difference is, you know, in a traditional uh, lead generation is happening, uh, through marketing efforts and through uh, prospecting and also on a AE prospecting uh, uh, in a PLG uh, we are focusing.

On those traditional ways to do lead generation as well. So it's not uh, in PLG marketing and SDRs are not generating leads. It is in addition to Uh, leads are also generated through free trials and free Uh, That might be your main product. It might be a sidecar product that is predominantly its only purpose um, to do lead generation.

And this particular difference between these two creates an opportunity to understand what PQ is are versus what MQL is are, and how we combine those two in one unified. Um, The second key difference is in a traditional model, we often have company information about the prospects that we're reaching out to.

We are using, you know, we build target account Uh, We have ZoomInfo we identify contacts that we want to reach out to in those target accounts. By the time we are reaching out to a prospect, whether that's through marketing or that's through sales, we have a sense of where that human being works in a PLG model.

In many cases, we don't have that information because users free people that are on free trials or free plans. They're using personal email addresses and I'm using a personal email address, does not let us know that they are using our product or service. Um, At Zoom, at Netflix and so on. Right. Um, and and that's where identity resolution comes into play.

The third key difference is that in a traditional model, when we are reaching out to prospects, we are customizing our messaging based on. Enrichment data that we've gotten from ZoomInfo and other providers and intent data that we may have gotten uh, our ABM platform, um, who is very familiar with your product and is an active user of your product, it's probably a really bad idea to tell them the five key things they need to do to get started to use your product.

Right? So product signals become an important part of the company. Um, The fourth big difference is in a traditional model, prospects have to be educated and informed about the value of the Um, But that, because they haven't actually used the product, right. They've seen your marketing materials.

They've maybe gone to your website, maybe. In a BLG model users already have a pretty deep understanding of your product because they're users of your product. So educating them on product is not necessarily a compelling way for an SDR to be successful in, in that role. And then second, last difference is in a traditional model.

Our prospects may or may not be technical. And in a PLG model, often the users that you are reaching out to have a much higher level of knowledge about your product, about the technology, about the space, and that really speaks to how SDR teams or teams that are doing outreach to users have to be retrained and trained in ways that we've never seen before in a traditional go-to-market market.

And then last thing is in a traditional model. Generally speaking deal sizes tend to be um, which basically means customer acquisition costs can be higher, right? Because if your deal size is 60 K, you can have an AA and an SDR spend 40 hours of their time acquiring that customer. Um, And so a human led approach ends up being totally fine from a unit economic standpoint.

In a PLG model, often deal sizes tend to start smaller and customer acquisition costs therefore have to be much lower, which then brings up the need for automation. And instead of a human led motion, a technology led motion with a human assist Um, uh, essentially how we think um, automation really changes and it's all driven by unit And then I will deep dive into these six key areas and workflows for each of them.

Merging PQLs and MQLs

discussed was how are we bringing PQS and MQL? What are they, how do we bring them together? And I think of it as three key steps. First and foremost, you need to take all your free customers and find a way to score them so that you can identify which ones of those represent the highest opportunity for conversion.

um, to do that, there are sort of three steps that you can think of The first step is identifying in your product what are the key aha moments that signal to you as a company that a user of your. product is getting value out of using your product So these aha moments really should be captured as what we call value metrics.

Um, At Dropbox, for instance, when a person shared a document with another human being that was considered a key aha moment or a value metric or a value moment. So first figuring out what are the five to ten value moments, and then defining those and turning those into metrics becomes important Second is you want to score every user based on usage of your product.

And when we think about usage, we want to think not just about snapshots. We want to think about the growth of usage and the consistency of usage. So for instance, what I mean by growth is if you are an HR recruitment platform, um, where, you know, recruiters can do searches and find candidates. Growth of usage would mean that the number of searches they are doing over time is increasing and consistency of usage would mean they are doing searches on a regular basis, whatever that regular basis is.

And then the last part of it is quota scoring, which is different from usage scoring and quota score is predominantly if you have pay wall gates right? Where I can only do five searches as a free user. um, in order for me to do a sixth search, I have buy right. Being able to see how fast a user is hitting that limit and how often that user is hitting that.

Uh, Becomes a different type of score, which is called a quota score. When you combine usage scoring and quota scoring, you are able to do PQL scoring effectively.

All right. So now that we've done, our PQL scoring and the output of that basically is a huge list of users that have all been scored on, let's say a point system, one to um, based on what aha moments they've had with our.

What is the growth and consistency of usage uh, our product and how often, um, how quickly are they hitting quota limits? We uh, PQL is identified and you can set a threshold at any level to say, okay, anything above this is considered a PQL anything below this is not.

bit. Now the next step is UL's with MQLs. Because often we find in a lot Um, right. Because at the end of the day, a lead is a lead. It doesn't matter how the lead was acquired. So the first thing that has to happen here is you need to have good Salesforce setup, Uh, so Salesforce's objects, which are contact lead opportunity. And so. Do not necessarily lend themselves well to companies that have a strong product led growth motion. So you need to create the concept, either decide that you will use the contact object to mean. Or we see many uh, use a, create a custom object called user, which they can then attach to a contact object or a tenant.

Similarly, Salesforce has a default object that is called generally be, There should only be one account for a company. However, in a lot of PLG companies. One company can have for a lot of PLG companies, their customers, one company can have many different types.

Across the company that are using our product separately. Right? So these tenants or service accounts also need to be created as custom object in Salesforce. So you can have a Salesforce account let's say for Netflix, and then you can have 10, 20, 30 service accounts for your product and all the teams at Netflix that are using your product independently

so that's sort of table stakes from a Salesforce setup standpoint. The second thing that has to happen here is users have to be matched to contacts. And what I mean by that is let's say a person is your product and their name uh, Liz at, and they've signed up as liz@gmail.com. You also acquired A list of contacts from ZoomInfo, uh, for a company Netflix, let's say, and there is a person there or a contact there that is called lizGallagher@netflix.com.

Being able to do identity resolution between lizGallagher@gmail.com and LizGallagher@netflix.com and being able to say these are the same human being. Therefore, this user is now tied to the Netflix Salesforce account in our Salesforce instance is a key part of merging PQLs and MQL. And then the last part of this merge is ICP scoring, because as you saw in PQL scoring, we had not put any filter on these users based on whether they are ideal customers or not.

And often. Doing ICP, sub scoring is not possible until you are able to do identity resolution and match a user of your. To a company because I CP criteria often exist at the company level. is it a mid-market company? Are they a technology company? Are they based in a certain region? And so do they have a certain employee size and whatnot?

Right. And so I CP scoring is where you are going to take your PQLs and your MQLs and then, or your all leads essentially. Which of these meet our criteria for our ideal customer profile. uh, that further sorts, that big list uh, um, two leads that are now showing high usage, hitting their quota limits or approaching quota limits quickly and consistently can be matched to an account that we care about.

And the account is a target account and an ideal customer profile. And then the last part of it is segmentation right now that you've done all of this. And like visually how to imagine this is you have a huge table and every row is a user and columns include things like. Where does this user work? What is their PQL usage score?

What is their quota score? What's their ICP score based on the company that they work uh, Now we want to be able to slice and dice this big list based on things like. How are some of them engaging with different marketing content, with different marketing Um, We wanted to be able to see which of these users are ideal for a sales assist motion based on the fact that they tried to pay with their credit card, but could not succeed.

Which of these users um, director and higher level titles uh, suggest that they are able to be economic buyers for us, which of these users are stuck on certain product features and would most benefit from product help or outreach uh, dimension. Right. And so sophisticated, fine grain segmentation becomes sort of the way.

Operationalize these amazing that you've been able to identify and unify.

I see a question from. Ash. Yes, Ash, I am in an island paradise right now. I'll show you actually, I'm sitting right in front of the beach. Um, So the question from Ash is most of the time, these data sets sit in different systems adopted by different sets of people across the enterprise. Are there triggers web hooks, streaming data capabilities?

That, yes. Great question. You know, um, there are two ways to connect this data together. You're absolutely Uh, The first way to connect this data is you use a solution uh, Fivetran and dump all this data into your warehouse. And uh, essentially uh, a sort of an intelligence layer on top with your data engineering team, and then use a solution um, Hightouch or Uh, To push, some subsets of this data into systems of action, like outreach, HubSpot Marketo, uh, Salesforce and whatnot. Um, That's one way to do. Another way to do it is to use something like Falcon, sorry, shameless plug. And what we're doing is under the hood. The exact same thing that I mentioned, which is we create a snowflake instance, use five train and dump all this data in do ID resolution on it, and then give you a segmentation experience.

And then under the hood, use high touch or Censuc to push that data out to the systems of action that you care about. So those are the two ways that we've seen that

Question: What are the value metrics parameters from an end user perspective?

so you know, How we recommend doing value metrics is you start as get your product marketing and sales team in a room and write down your Uh, And, and I, I will get to how you get to this from a user perspective. You should first write down your hypotheses about what you believe the core value of your product is to your end. Customers ask yourself three levels of why, because often when I ask people this question, the first answer is a feature or capability and pardon my language, but no user gives a fuck about a feature or capability.

They're looking for a uh, that they, that they want. Right. So as an example at um, the first assumption we made was. Um, Dropbox is valuable when uh, create a document. No, it's not. They, the document only becomes valuable if it is synched to another device or if it is shared with another human being, because you are trying to collaborate with someone.

So our first answer for what is a value metric for Dropbox was number of documents created. Wrong. It ended up being number of documents, shared, and number of documents, synched across devices. So validating. So we come up with our list of hypotheses. Then we go and dive in our data to see, is there a correlation between our hypotheses and actual conversion events for users?

Right. So if I tell you that 90% of your customers that have not converted. Created a document. And 90% of customers that have converted also created a document. Clearly document creation was not the difference that made the difference between these two groups, right? Versus if I told you the number 90% of users that did not convert, never shared a document and did not have more than one uh, using our product versus.

70% of customers that converted, shared multiple documents and uh, two or more devices. Now you can see the difference. That's making the difference. This is where a lot of data scientists will jump on me and say, correlation is not causation. Totally get it. That's why you need a human. This is a human in the loop solution.

So you see the best causal inference engine is the human brain. Correlation will help you whittle down the hypotheses from 15 to five, right? And then you can decide as the human, ah, this causes this, therefore this is the value metric. I will also say doing that is in parallel to doing qualitative research and finding your 10 most engaged users and going uh, on how they use the.

What aspects of the product they use, what they love about your product? My one caution on qualitative research um, and I've seen product managers screw this up over and over again. I'm a product manager as well. Uh, We uh, qualitative research with the questions that we ask. And so if you're going to go down the qualitative research.

I would strongly recommend an observational approach rather than a Q and A approach, watch people as they use the product, as opposed to asking Um, Because I have not seen a lot of well-crafted questions that don't bias the user to give you the answer. That is your hypothesis anyway. Right? So in a perfect world, both of these go together.

But, uh, I said, on the data side, you want to be analytically. Come up with hypotheses, validate them with data. Uh, In qualitative research, you also want to be rigorous and bias for observational insights versus Cool.

Identity resolution

Now I'm going to get into a fun thing ID resolution. So when I was at Amperity, we built ID resolution for consumer companies. And so this is a uh, very close to my heart. So in a BLG motion, we have to think about identity resolution as three distinct problems. Problem. Number one is user to contact.

And this is the one that I was just describing earlier, which is. You have a million users and half of them are using personal email addresses to use your product. How do you figure out where, like, which contacts they map to in your contact list in Salesforce? Right. So I have a list of users here and I have a list of contacts here, and I want to be able to match these, to figure out, oh, this user here is actually the same as this contact here.

If I don't do that, I'm going to. A lot of duplications across these two sets and from a customer stand point that is problematic because as a user now I'm potentially getting two emails from you, two outreaches from you that are saying different things, but I'm the same person. I just happen to use a Gmail address to use your service.

And then you got my contact information on zoom info with uh, work, add a work email. Uh, And now you're bombarding me with, uh, the wrong message on two channels. Right? So the way to do user to contact um, there's a very literal version of it, which is named matching. It's easy. The pros of this are it's easy to set up.

Um, Same thing for email matching. You can do an exact match on name and email. Exact matches will result in a lower match rate um, one, a lot of times people don't give you their accurate first name or last name. Second, you end up with typos all the time. And so you will just get lower match rate than you expect.

However, the pros of this are easy to set up The matches that you will get will be a hundred percent accurate. Right? Then we have a more sophisticated approach, which is the thing that we built at Amperity. My previous company, where we do heuristic matching and the heuristic matching. What it's doing is it's looking at things like rarity of first name and last name.

Like Mona Akmal is a rare name in the United. Therefore, if you find a MonaAkmal at gmail.com and you find a Monaakmal@falcon.ai, the likelihood that these are the same person significantly higher than if my name was, I don't know, John Smiths, I don't actually know very many John Smiths, but you get my drift.

Right. Um, And other way to do heuristic matching is you look at the username in the email address that. And you try to see if there is some heuristic of first name, last name, because a lot of people have their email addresses, be a reflection of some combination of their first name and last name. So you can get really, really fancy.

What that does is it improves your match But it becomes more and more complex to set up uh, more and more complex to maintain. The second type of identity resolution that we have to think about is user to account, which I also alluded to earlier in user to account, you are trying to say Monaakmal@gmail.com actually works at Falcon AI, which is a Salesforce account that I.

But that, that is a target account that I want to pursue. Um, With several of our customers, we've been able to identify a whole bunch uh, personal email addresses of users and map them to accounts and help them see, holy shit. Like in this account, we have a lot of people using our product already. So that completely changes your perception of how much you've penetrated a target account and what tactics to take to turn that into an enterprise bank account.

Right. And in user to account the way we think about it is you have to use third party data enrichment. So you can use like a vendor like people or um, to enrich your user information. And try to get more metadata on your uh, and where they might be working. You can also do things like domain matching.

So for instance, if someone has given you, like, let's say somebody works at, um, Adobe, uh, but they've told you their name is Simon@adobe.co.uk and you want to, uh, and you have a person that uh, jane@adobe.com and you have an other person that works in some other subsidiary of Adobe being able to do domain matching in a, again, there's an exact way.

And there's a heuristic. You can use data sets like Crunchbase, for instance, to say adobe.co.uk and adobe.com and Adobe uh, whatever XXX are, all one company. And that company is Adobe. So that's how I'm going to match users to accounts. And then we have tenants to accounts, right? Uh, This was actually, we're talking to the head of BI at, uh, Mixpanel

and they were telling us that, you know, they had to go through a massive, uh, data cleaning project um, they had so many duplicate Salesforce accounts because for every service account they created, they would create a Salesforce account. So now, if we have a hundred service accounts with different teams in Netflix, we now have essentially a hundred Netflix salesforce accounts that represent actually one company, not a hundred different companies. Right. So that's sort of also where we use similar techniques, heuristic versus um, and use third party data sets like Crunchbase to say, no, these are actually the same company. One other thing that I would caution is if you don't have the um, product instrumentation set up.

So like if you're using amplitude or Mixpanel or Pendo or Google. Make sure that your engineering team, when a new service account is created or a new tendency is created, that they are able to, uh, put in the name of the uh, which will then help you attach that to a, a Salesforce uh, in a much easier way.

And then second thing is, make sure your Salesforce structure is set up. And again, What I recommend is creating a custom object for every service account, and then being able to join that service account with a Salesforce account using account IDs, either way.

Product signals

Now let's move to the fun stuff, which is product signals, right? So we are trying to identify who are our best users and who should we be going after to convert them from free to paid? And so there are three things to consider here. One is snapshot data, which is you identify that a customer is using key features of your product.

And this is like, you know, in the maturity arc of PLG, this is where companies generally start. Um, So. A customer, let's say we're back to the recruiting example. If a customer, if a user is doing lots of searches, they are downloading a lot of resumes. They are sharing those resumes with hiring managers as a recruiting platform.

Those are all key features of the platform that suggest um, high engagement and high value. Um, You can do snapshots of these to see. At any given point, which user has used feature a, B, C, D, E, and decide what to do with that based. So you can use it in a couple of ways. You can use that for segmentation.

You can use that for personalized outreach, like sending a marketing email to someone saying, Hey, I noticed you shared your first document, or I noticed that you shared your first resume with a hiring manager. Did you know that if you did this. Insert new feature. Um, You'd be able to save a lot of time and manual effort, right.

Then we get uh, stuff that's a little bit more sophisticated, which is okay. So if I have a snapshot that tells me a user is using this feature, this feature, and I have 10 features becomes a lot of information to process. So now you can create higher level segments. For users that are using these four features, I consider them an advanced user for users that are using these three other features.

I consider them an intermediary user. And I'll give you Uh, One of our customers seek out one of their key indicators for an advanced or sophisticated user. And seek out is a recruitment platform, right? It's used by recruiters to find candidates and then share them with hiring managers and so on.

One of their sophistication features is if a recruiter saves a search and uses the carrot in their uh, experience, it indicates that they are an advanced user. So you can come up with some pretty simple logic to group users as advanced. Intermediate novices that helps you make sense of the 15, 20, 30 possible things that they can do in your product, because it's very hard to look at 30 numbers and try to make sense of it.

Right. And then the next thing that this is where, to me, if you are exploring your PLG motion as a, as a, uh, new, it's a new experience for you, you're probably thinking about snapshots. If you are a little bit more farther along in your journey, you're thinking about growth rates, because we don't just care about the fact that someone has used a feature.

We care about how that usage is changing over time. Is it getting more entrenched or less entrenched? Right. And is user sophistication changing over time? Are users actually going from being novices to being intermediate, to being advanced? Is the uh, usage across all users and all tenants within a company increasing over time, right?

Those growth rates require you to have a historical perspective instead of just a snapshot point in time. This is the world as of today. And we think growth rates actually unlock deeper, more relevant insights that are less noisy than snapshots. And then the third element of product signals to consider is consistency of usage.

Right? So if I told you that there was a person that started using your product and they used feature a once. Then they used it 10 times and then they used it once and then they disappear. How does that compare with someone that uses your product consistently four times a day, intuitively you can see that these are pretty different profiles and they need to be approached very differently.

And in all at the end of the day, the point of this entire workshop, You need to understand your user behavior at a significantly deeper level to do personalized outreach. Consumer companies are very good at this. B2B companies are just starting to figure out what this is, right. And so consistency of usage is really how is a given user using your product or a key capability in your product over time?

Is it a flat line is very spiky? Was it a one-time they tried it and then they disappeared. Right. Which actually happens a lot in free trials and ends up with bad PQL signals is you'll see someone just, you know, spike and hit their limit and use a bunch of things. Value moments have been triggered, Uh, They are a PQL. No, they're not. You want to look for consistency of usage over a period of time to understand if they were just testing or they were actually

To me, this is squarely in the realm of using a tool like amplitude Mixpanel Pendo, aggressively and like in a very sophisticated way to figure out where are end users dropping off in your product usage experiences and look at it as a user Amplitude does a great job of showing you user journeys over time, all events that are, that are associated with a user linearly laid out over time, and you can see where the drop-offs are happening. Um, Often I don't see that um, a top of mind issue for go to market teams, right?

Go to that. To me, is squarely in the realm of product management using a product analytics tool like. Um, To understand customer behavior and user Uh, So this is where we go back to qualitative again. Right? I think um, interviewing, uh, users that were using your product in a highly engaged way and then turned out and stop using stopped using is a great way to get signal on where the pain points are, where there's lack of value.

And you don't need to talk to a hundred to figure out what the patterns are. You need to talk to And so I, I would encourage uh, you know, once a quarter, at least you should be looking at your churned customer base and having deep conversations, give them uh, reward and incentive, obviously to spend 30 minutes or 15 minutes with.

Um, But it is a good way to match what you will find in amplitude, which is data at scale. And then also qualitative research with people that have recently turned out.

SDR role in PLG

So what does the SDR do in this new world? I really believe the SDR role is going to get completely transformed with product led growth first.

Support becomes a very important part of what you're doing in sales development, because in order for customers to become or users to become worthy or worthwhile leads for us to pursue, we first have to make them successful. And in order to make them successful, we have to support them in the places that they are stuck.

Um, That can be done. Through a great product um, automated hyper-personalized set of campaigns to onboard customers and get them stuck. Um, It can also be done through rewards. So we've seen some customers create usage. Like congratulate your users and reinforce positive behavior when they use your product or a key capability consistently over a period of time, because you're trying to help people build habit right?

Half the time. The reason why people don't get value out of your product is not because your product is not valuable. It's because they have a different workflow and a habit of doing things. Even though that way of doing things is really painful. They have a way of doing things. And they have a habit built around it.

This has been something that has existed in game development for a long time. How do you create positive feedback loops that encourage habit forming? Right when I was at Zulily, we talked about this a lot. How do we make Zulily, a daily habit for a young uh, who was our ideal customer? The second aspect of the SDR role that I think is, is very interesting, is to move into a sales assist role, right?

Which is I'm a user, I'm a developer and a lot of BLG companies do tend to bias towards a technical audience. These are people that are very sophisticated users of your product. They often have no idea how procurement works, how privacy and compliance works, how to build business justifications for purchasing software, or like how to even purchase things online in a professional context.

So an SDRs role, uh, evolves, to support that user in logistics of making purchases, like going through a purchase. Helping them navigate the procurement process because they're unfamiliar with it and helping them build a business justification that will be relevant for their stakeholders internally, to make the buying decision.

Right. And that's generally stuff that, that SDRs have not done in the past. And then the last part is I believe SDRs in this world can actually do a lot of. Gary quota and close deals or move them to AEs. So for smaller accounts, and we see that, you know, we were talking to someone at Twilio and their SDR team is a quota carrying team because for a lot of smaller accounts, you're trying to go from free to paid.

You don't need an AE to be involved for that. An SDR can drive all the way to close in that scenario.


SDR training

So if, if the SDR role is evolving, then what training should support that transition and that. First is product training and in product training, you have to really, really go deeper, right? It's not as much. Yes. It's definitely still about being able to communicate value succinctly being able to discover pain points and whatnot equally important now is having a pretty deep knowledge of the product and being able to answer level one.

Um, Questions from a technical depth standpoint, right? That's a gap in training with SDR teams right now that I think needs to be addressed. Second, have a deeper knowledge of your competitive landscape um, again, your users in this case are very knowledgeable about the space. They probably know your product better than you do because they actually use it and it was built for them not, you.

So to have a knowledgeable conversation that is high value for this type uh, engagement, you need to understand the competitive landscape, your differentiators, uh, uh, a good technical understanding of your product and where it fits in and be able to do uh, level um, conversation around pricing again, with smaller deal cycles, uh, talking to a more technical.

They don't want to have a large production talk to an AE and get, uh, custom pricing built out. They want something that is a little bit more lightweight, a little bit more transparent. Um, So having all that information at your fingertips becomes really important. And then the last part of it is procurement, which is helping the, the person that you are reaching out to the MQL or PQL that you're reaching out.

Helping them quickly understand how you can help them with privacy compliance, business justification creation, and what a good procurement process looks like so that they know what they need to do. They love your product, they want to purchase it. They just don't know how to in their company and your job is to really help them understand how procurement works and turn them into a champion that is going to go make things happen


and then the last thing is around automation, which we talked about just a reminder, Why is automation important? Because PLG companies generally tend to start with lower deal sizes, which means customer acquisition cannot be heavyweight human led and costly because the unit economics on that are not great.

And so. Whenever, generally we want to go for cheaper customer acquisition, uh, an effective path to do that is automation. Right? And so what are the types of things that we can automate? Uh, And how do we automate. First and foremost, you need to have a lot of attributes associated with the contacts that you are going to reach out Um, Those attributes need to include things like what product features are they using? What value moments have they had already? What is the growth rate of their usage? What is the consistency of their usage? Are they sophisticated? Advanced? Uh, Intermediate or beginner, are they stuck Um, Have they actually tried to purchase, have they tried to contact sales? Have they had a support interaction with your company? So being able to get all of those attributes, you need to almost think that contact object in Salesforce is a hundred X richer in terms of attributes than it's ever been before.

And it's live, it's able to show. Usage and trends over time, not just a snapshot of where they are. Snapshots are great at things like this contact's name is Mona because it's not going to change over time or their title is blah because that's also not going to change a lot over time. Right. But when we're thinking about live attributes, like usage, like sophistication, Those are things that do and should change over time.

So you need to think about those as live attributes, that show you trends over time. Not just a snapshot. Once we have this very rich contact row in Salesforce, we can then do sophisticated segmentation and sophisticated segmentation is key to automation, Um, Because what do we lose when we automate?

Generally we lose that human touch, that personalization. Wow. If you have enough data and you create enough refined segments, you don't have to lose out on that. Very personalized human touch. You are just using technology to imitate the human touch instead uh, you know, manually doing that work yourself.

Right. And so fine grain segmentation based on these attributes that we've talked about becomes really important. And then the ability to sync these segments. Two systems of action, like outreach, HubSpot Salesforce Marketo becomes really important because that's where you're going to do something with it.

Right. And then last is even automate the outreach, right? A hundred times the outreach sequences that you have right now, and auto-enrolled segments into those sequences. And those can be outreach sequences, Marketo sequences, HubSpot sequences, whatever marketing automation platform you use, whatever sales automation, enablement platform you use.

Being able to push those segments and then just auto-enroll them. So that the first outreach doesn't even have to be a person. It can just be a highly personalized email that is being sent to the right user at the right time with the right message. Except no human was involved. We just set it up and then.

We let the system just run this right? Only when people respond back, should a human get involved. That is going to save you thousands of hours in terms of human time, which then dramatically brings down your customer acquisition costs to take your free customers and turn them into paid customers. And, obviously you, uh, last point on this is in order to create that many sequences you need to hire and have bandwidth within your marketing organization, um, to be able to pull that off, right.


Mona. Thank you so much. It's a pleasure to make your acquaintance, uh, Samuel and uh, exchanging ideas for the last 10 days or so. And I have been thinking quite a bit about this transition. Since many SDRs are trained to do the exact opposite of the kinds of things you've suggested and the.


Person who will lead this feigning does not exist inside of a company. If you kind of step the highest level, you've got a lot of growth marketers. You've got an emphasis on digital automation platform to do onboarding. You've got PQL platforms, but you really don't have somebody that sits in the middle there that says, how do we engage our SDRs to level them up, to listen, to guide and to scout.

I am so aligned with you, Chris. And you know, the fun thing is that creates an opportunity. I think there's a big business to be built in leveling up and retooling and retraining SDRs at all these PLG companies so that they can stay relevant in this new, um, new age. Right. And you're totally right. It doesn't.

And so I think it's important for everybody in the community to start defining the skills and abilities, uh, and to understand at what point you're going to create a cross-functional team that pulls from say support. Yeah, uh, that pulls a bit from marketing that pulls from customer success to try and reinvent the role and move away from essentially forcing an outbound motion into a product, a understanding type of arrangement.

Totally. That's brilliant. I love that suggestion and yeah, I think the support angle is really, really fascinating because. Often,