[Music] [Corey Bayless] So today, we are going to crash course all of you into RCM and how you can empower it with a native Marketo integration. So not really using much outside technology, if you have Marketo, this is specific to Marketo and how it works. So our agenda essentially is going to be strategy, the different ways to look at lead lifecycle, implementation, gotchas, launch, optimization, and then finally, we're going to wrap it up with our takeaways. So let's get started. I'm going to first start out with introducing Jenny Robertson. [Jenny Robertson] Hi. So who struggled getting up this morning making a 9am session? Good. Good.

We brought a lot of goodies for everybody.

We knew it was going to be hard to make it. We appreciate you getting out of bed. So I'm Jenny Robertson. The reason I'm here today, I've got over 20 years in technology experience. I know those in the back are probably like, "Wow, she looks way too young to have 20 years experience." Those in the front, they see the wrinkles and the gray hairs.

But I've been on the marketing technology side of things for the past, I'd say, 15-ish years probably. I've been in Marketo for almost 14. I've been a 10-time Marketo champion. I lead the Atlanta Marketo User Group for a long time. I don't even know how long, but I'm just really passionate about what I do in the marketing community. So that's why I'm here today. A little bit about me outside of work. I have a big family. I'm a really proud boy mom. I have two boys, six and nine. Any other boy moms out there? All right. All right.

And then outside of work, I hike and run.

So on to you.

So that's Jenny. I think she fights crime in the side projects, but it's all good. So I'm going to do my best to follow that up.

So a little bit about me, just, kind of, the key points. I'm a 10-year veteran in Marketo. My blood is definitely more of a maroon, purple, red color depending on the acquisition and where all you came on board. But definitely started back when Marketo was Marketo and then the acquisition with Adobe. So been there ever since. Couple of key things about me. I've worked on some of the biggest instances in the world. So Microsoft and AWS. I didn't choose this life. The life chose me. Just throwing that out there. Administrator for many, many, many different Marketo Engage instances and many different clients. And then a couple of interesting things about me also, I'm a nine-year veteran, SDE coder, then Python. So I got a pretty heavy technical background as well, which has also really helped steward my career. And so it's, kind of, turned me into, sort of, a data engineer, right? And well, that's important when you're in Marketo because you need to control that data. So that's what we're going to talk to you about today, right? How do we control that data? How do we make it sing? How do we bring everything together? So data compliance, GDPR enthusiast, all the various things, right? We like to keep it safe while keeping it real.

All right. So about us. Why I believe Jenny and I are on this stage. We have a combined 24 years of experience. So what does that mean? Well, it means that you've seen stuff. You've been in the trenches a long time, right? It's grit. It's pain. It's clicking a mouse a lot. It's all of the creativity that you can to try to solve your problems in a pragmatic way that leads to efficiency, but when all done correctly, the system essentially prints you money. Like, it's like a crank and you just, you feed that crank and it just spits out money on the end of it, which is pretty cool. So 14 years combined as in Champion program. I mean, Jenny's got 11 of them.

So yeah, Certifications Galore. I cover more of the cloud side, but Jenny's CRM, certified admin, Adobe certified master, just absolute amazing accolades and one of the best people I've ever gotten the opportunity to work with. So really excited to be partnered with her on this.

Okay. Let's get started. So let's talk strategy, right? So what is RCM? What is the Revenue Cycle Model? And I just want to throw this caveat out to everybody that this is going to be a brain dump and knowledge transfer session. So we're going for substance and delivery. So if you don't think our slides are the utmost beautiful, well, that's okay because we want to deliver the goods. That's the key here, right? So RCM is a visual representation of the marketing and sales funnel which shows you how leads can move through the system, right? So if you have a large database of people, RCM will help you put substance to those customers. Now what do I mean by that? Well, I have the ability to talk to my data in a way that allows for me to see the movement all the way from anonymous, known, they become a person, and then I can communicate with sales.

It helps you understand the customer sales funnel. It creates a visualization of that movement in a way that's you can understand and comprehend and pass off to leadership so that they understand what's going on. With the general concepts of, you have your first-touch capture, you've got your multi-touch attribution and then you have your last-touch which leads to opportunity. That's the general forecasted method of way of doing things. Now with AI, Marketo Measure, there's a lot of new innovations that are coming that I would highly recommend, right? This is a little bit of a linear way of thinking about the sales funnel and things are adapting and so do we, right? So this isn't going to solve every single problem we have, but it's really going to give you an idea of what do we need to do to get from A to Z.

All right, so what are the RCM stages? RCM stages typically consist of an anonymous record. What's an anonymous record in Marketo? An anonymous record in Marketo is somebody we don't know. They come to a landing page. They take some action. What Marketo is doing in the backend is they're actually recording all of that activity as it happens, right? As soon as that person becomes known, they have an email. Their primary index in Marketo is an email, right? All of that activity data that was being tracked then becomes something. It becomes a searchable lead record in the system, right? Okay, so moving forward. So now we know who this person is. Now what? So we can mark them as an inquiry or known person. Email is something, right? Email is something. So a really cool thing that you can do, which was something that Marketo just launched a couple of years ago was reCAPTCHA, right? Now if you're going to qualify data, you want to make sure that it's not a bot, right? Bots are actually one of the biggest issues that market automation experts face. Believe it or not, this is like a real thing. So CAPTCHA essentially allows us to say, "Okay, anybody who is a score of above 0.5 or greater than 0.5 generally is a real person." Now you might be asking, how does reCAPTCHA do that? Great question. So reCAPTCHA is essentially a component that works on the backend on your web page that has a forum on it, and it is looking for lateral and vertical movements from somebody who's going to be looking at your page. So a human, when they move their mouse is going to be all over the place. There's no structure to it, but a bot moves vertical and horizontal. ReCAPTCHA can look at that and they see that, and they know that. That's how it works and so that's how it knows if you're a bot or if you're a person. So kind of a fun little fact. Okay, so moving forward. So after we qualify, we can disqualify all our bots, keep all of the bad data out of the system. Let's always do that, right? You have now your engaged section of, they're doing something, right? They've interacted with something. They've become engaged, right? I'm getting them to a point where now I can actually start to qualify my data. That's where I basically, and you can build this however you want, but I like to basically say, "Okay, this person is an email. They have some form of engagement." Things are happening, right? Great. Cool. Awesome, right? So now what's next? Now I want to turn them into a marketing accepted lead. So that means they've subscribed.

That means that they're not a bot.

Excuse me, and that means also that they have a profile. So general profiles can be set up however you want. Whatever the business model entails, right? Email's not none. First name is not none. Last name is not none. Industry is not none. Job title, right? If you've got metadata on the backend for that record, you can create be as creative as you want. But you essentially want to create good looking profiles that have good data that lead to cleanliness, right? Data in, good data in, good data out. Garbage in, garbage out. So if you control all of those vectors, you're going to be in a really good spot. So let's get into what is an MQL. So an MQL is a Marketing Qualified Lead, everybody's favorite term, right? A lot of different ways to do it, but one of the biggest ways you can also, kind of, see if there's a lot of movement is through lead score. Lead score is going to be your determination of did somebody go above a threshold of 62, or 100, or 150, or however you have it set up within your system. The general idea is, is that if they have some form of a lead score, then I am going to be able to qualify them and say, "Hey, sales, this person is interesting." This is somebody who I qualified and said, "You know what, they're not a bot, they're subscribed, they want to talk to us, all of the signals are kind of there that I want to listen to." Now do I think lead score is the end all be all, at way of tracking and pushing people through? I don't, but I'll be honest with you, I'd rather have one than not.

All right, now we're moving into the sales side. So sales accepted lead and SQL, right? Sales Qualified Lead, Sales Accepted Lead. Now what does that mean? So if you do it right and you have a CRM with a sales team on the backend, I want to make sure that my system is in sync with CRM, with Salesforce. I want to basically say, if there's a note that comes from sales about communication out to this customer, I want to make sure that I have those details in Marketo. Why? Because if I have a standardized process of collecting this information, then that means I can build triggered campaigns and batch campaigns on the backend to move them into their various sections in Marketo, right? Everything is always moving. The more hands-off on data I am, the more valuable it becomes, right? If I can preserve my time through automation, then that means I can be really productive with how I'm able to produce and I'm able to target my customers. I'm able to email them correctly. I'm able to get things done, right? Big things. All right. And then everyone's favorite. SQL, Sales Qualified Lead that leads into the coveted opportunity. Some form of a dollar amount that's associated with this customer and then moving forward, if all things line up and the stars align, we win the contract, right? The path to success. That's RCM in a nutshell.

So lead lifecycle considerations. Collaboration with sales is absolutely critical, right? You keep in mind the general perspective in marketing is, how do you want to be communicated with? Do you want to have a barrage of emails coming into your inbox or do you want it to be pragmatic? Do you want it to be strategic, right? If somebody's talking to you, I want to have something that's relevant to me at the time. I don't want random things or people reaching out saying, "Hey, you should go to this auto show because you're such a great person." And I'm like, "I don't know anything about. I'm not a mechanic. Why did I get this email? How did you get my email? What is happening? What is this world?" Right? So you want to be really want to be specific about how you communicate with your customer, especially your subscribed customers. It's the one thing that you have to protect because it once they unsubscribe, they're gone. You can't talk to that customer anymore. So you want to be very specific with how and why and when you talk to your customer. And it should be based off of signals that you're getting.

So these four stages in my book are essentially the most important because MAL, MQL, SAL, SQL is going to be my conversation from qualification all the way into what is sales going to do, right? Can I only get sales the best data? And if I can, then I'm in a really good position.

So a couple of things when you're creating MQLs and why they're important, right? An MQL, like I said, qualified lead. Keys to making sure that it's the right data. This person that you're reaching out to knows what the company is. They know the brand. They've interacted with something. Never qualify an MQL on an open. Ever. In fact, almost don't ever qualify an MQL on a click. Because those are interesting signals, but you really want more. You want more substance because everyone's getting emails, right? It's a tough practice. It's not as easy as just sending somebody an email and saying, "Ah, yes. Now this person's qualified," right? Listen to the signals. If they're not interacting and they're becoming stale data, remove them.

Maintain that subscription. Don't over send. Even the best emails, most personalized emails will get unsubscribed. But as long as you can keep that threshold below 0.0%, then you're in the right spot. With anywhere from 89% to like 95% deliverability rate, that's a really good spot, right? Deliverability means money.

That's what happens, especially at scale. Super, super important, right? Clicks, fills out forms, downloads white papers, attends events, attends the webinar. All of these things are positive signals that are interesting to sales. And guess what, I want to collect all of that data.

All right, so let's talk about person scoring best practices. So a lot of times, there's ways to do this but everyone talks about documentation. Why is documentation important? Well, documentation is important because I'm doing everything I can to remove this single point of failure. What does that mean? So when I talk to people in my org, I want to make sure that they understand how this system functions. I want them to be able to have access to this documentation and this is the key that I have in Marketo, and I want this to match what's in Marketo. If my scoring changes, I want to change the documentation, and then I want to change it in Marketo, right? Create a CSV, list of actions, all of the different things, capture your demographics, capture your person scoring and also your behavior scoring, fills out form +5, however you want to do it, right? Work with sales, figure out what do you think is a good buying? What are good buying signals, essentially? So person scoring types, right? We got Demographic Score. So Demographic Score combined with your activity score is super important because if I have somebody who's in middle management and I'm in a buying group, I want to make sure that I'm talking to this person, but I also want to get to know all of the other people that are around that person, right? The influencer. The influencer is who I want to talk to. And I want to make sure that they're listening 'cause if they're not, I don't want to bother them, right? I want to be really practical with how I'm doing things. Needle in a haystack, right? Kind of how you want to think about it. But when I have market automation, instead of a one-to-one reach out, I'm able to send 30,000 emails out. And if 100 come through the door, that's pretty successful, right? We're starting conversations. We're building relationships. We're being kind. We're listening to our customers and we're being customer obsessed.

Couple of different scoring options also, with negative scoring. You gotto attend webinars, missed webinars, attend company events, missed the company event, and so forth, right? Then you also have negative scoring. When something bad happens, take them down a few notches, right? Or they go still, take them down a few notches, right? You want to just, kind of, keep it as, sort of, a functional way of keeping everything in line.

So the power of the Person Score and why it's important. So if I'm able to qualify data and I'm able to remove bots and I'm able to push people over to sales and I'm very specific with who I send to sales, then I'm really, I'm taking advantage of their time, right? So a couple different ways to look at it, you can tier your data, you have a tier one, you got tier B, you got tier C, and so forth, right? Anybody who's in 20, you can also use this for your targeting. It's a great way to say, who's top of funnel? Who's middle of funnel? Okay, who's getting ready for an opportunity? Who can I close on, right? My job is get sales the best data. So if we go into looking at the Marketo side of it, I've got anonymous, known, engaged, MAL, MQL. This is my Marketo side. Now you can sync data with sales at any time, but I only want to sync the best data with sales. I only want them focused on the best people. I don't want them to have just a myriad of all of these different people and only some of them are qualified. I want to be sending them the best. Everyone's time is really expensive, and you need to think in those terms. So SLA is an interesting one.

Essentially, an SLA doesn't actually work, as well as you might think because it doesn't take into account weekends, but we'll move on from that. It's fine. It still works. It's nice. But you can also do it in a different way, okay? But essentially, kind of, that's your general flow for your success path. Now RCM isn't linear, it's actually circular, but we'll get to that. And then here is your sales cycle. So now every time my sales team talks to a customer, I want to have that automated. I want them to add the notes. I want them to have a pick list of different values that they can choose of how did that conversation go? And when they tell me how that conversation goes, it stored its metadata on their record. So I can take advantage of it, right? Positive metadata is how we automate the sales cycle.

And as you can see also, how do I do that? Well, I have a person status. If I have all of my pick lists determined and sales understands that and it's clear documentation, I can automate it.

All right, so let's do a couple of things. Different ways to look at lead lifecycle. So here is the power of the 360-degree journey. So I disqualify data, I can then recycle them into a nurture stream, right? Or I can disqualify them and remove them from the flow. Be very, very specific with the data and the people that I'm talking to. At scale, super important because I don't have time to go through 50,000 records. I just don't, right? So I just need to do the best I can to automate the most that I can to have the least amount of touch on my data and let sales be that manual integration. If I can automate my job, it makes it a lot easier for sales. So here's a bunch of different flavors of RCM that you can look at. There's all kinds of different ways you can build it. I built the top one just for this presentation alone and created the entire flow, did all of the different things that you need to do and all of the different qualifications. But you can see I have MQL, SAL, and SQL with rejected. Don't have budget. Timing's not right. All of those various things. Listen to those signals, right? Disqualify that person. If they're subscribed and they still want to talk to us review that record. What are they doing, right? Do they have a good lead score? Are they stale data? All of these things are available, right? If you understand data, then you're able to pull these different metrics.

And then you can, to track these things with proxy fields or Success Path Analyzer or SLAs out of touch, all of these various things. You can use Revenue Explorer Report. It will tell you the SLAs that are missed based off of the thresholds that you put in place, right? So there's a couple of different ways to track them. Jenny's going to definitely go into some proxy fields and how I think it's a lot better to use Smart Lists to track them, but we'll get to that. All right. Then some open questions that you can ask sales to get started.

What is the ideal lead? Who do you enjoy calling on? What's a functional area person a that leads to a lot of success? Really, it's about working backwards from sales all the way into marketing. Those are, kind of, the keys to success. All right, so that's a lot about strategy and now I'm going to pass it over to Jenny who's going to talk about implementation. All right. Thank you.

Okay, so we're going to cover for implementation, some ways to think about implementing efficiently and scalably and following best practices. And the part I'm most excited about is to share a bunch of gotchas you need to watch out for when you're implementing. But why is all of that important? If you don't have an efficient and scalable lead scoring and Revenue Cycle Model set up, you can lose QLs. You could have false QLs. It's easy to lose sales trust. It's a lot harder to get back.

So we'll start talking about implementing person scoring. It's a best practice to make sure you have your scoring in centralized programs versus just throwing scoring in a bunch of different smart campaigns throughout your database. It's just an easy way to be able to find your scoring, optimize your scoring, know how you're scoring. It's really a best practice to follow. We had a client once that was throwing in +20, +30 all over the place for all of these event flows, and then they couldn't figure out how people were QL-ing, and they had some marketers scoring wrong. So if you have it centralized, it really makes sure your system's running efficiently. You should also always use scoring tokens. If you're using scoring tokens when you're ready to optimize scoring, you have one place to go and update the score. You don't have to go find all of the smart campaigns that you're scoring +24. You can just go update the tokens. So tokens really helps you efficiently optimize your scoring as you change things going forward.

Also, only score high value demographics. If you're scoring 20 different demographic fields, and then let's say you have a list upload, maybe you had a webinar, and it wasn't integrated. If you upload a list and it's a bunch of people, say 5,000 people, and they need to run through 20 demographic scoring campaigns, that's going to backlog your database for a little bit. So when you're scoring demographics, really think about what the key demographics are. You really don't need to score for all of them.

Also, to prevent strain on your system, make sure you don't add unnecessary wait steps. Wait steps deprioritize smart campaigns. So if you don't need them, try not to use them. Try not to request another campaign unless you have to. There's definitely use cases for it, but you don't want to do that everywhere. And then don't reference another smart list within a smart list. Nested smart lists can really wreak havoc.

And some general best practices for scoring. Make sure sales is aligned to what you're considering a QL, how you're scoring to get there.

Corey touched on that a little bit earlier. Keep an updated requirements document. You really should not be implementing scoring or changing scoring until your scoring doc, like Corey showed earlier, is updated. If you just go in and update scoring and say, "I'm just going to update the requirements doc later. "Guess what? It's not going to get updated.

Align your scoring to your Revenue Cycle Model. They are not independent. They're definitely related, so make sure you're thinking about both when you make changes. Make sure you have processes for recycled and rejected leads. It's not just turning somebody back a stage. You might need to change the score. You might need to send alerts. Who knows? Make sure you're thinking through those processes.

Use consistent, easy to understand naming conventions. For us, an example I can give you, we score events different levels. Some events get scored high, some events get scored low. It depends on the event type and the content of the event. We use naming conventions in our programs so that we can easily scale how we score each event instead of having to put in a special token in the event itself. And then always test, test, test. I can't speak enough about testing. It's so important.

So we'll talk about a little bit about implementing RCM. So you saw those pretty pictures Corey had earlier that show the Revenue Cycle Model with all of the nice arrows. When you're deciding how to implement RCM, you can set up transitions inside all of those little arrows, or you can use a program in Smart Campaigns. I definitely prefer using a program in Smart Campaigns. If you're setting your transitions up inside the Revenue Cycle Model, all you can do is change the stage. But if you set it up in a program, you could take other actions. Like, for us, for example, if you have a QL, you're going to update the revenue stage. You're going to update a status. You're going to send an alert. You're going to change an owner. You can't do all of that within the Revenue Cycle Modeler itself, so it's a really great practice just to have a program where you manage all of that in one place. Just make sure if you do it in a program, you're referencing the stages in your Smart Campaigns because it's not going to know that this score should make a prospect go to engage. You have to include extra filters for that.

Also, you can use your revenue cycle program to know when to sync records to Salesforce or Microsoft Dynamics.

We had a client once that had over 100 different places. They had the sync to SFDC flow step, and it ended up really causing a lot of problems. They had records thinking they didn't want to sync. They didn't know which campaigns were doing it. Having a centralized place where you have people syncing to Salesforce really helps, down the road. For us, it's in our revenue cycle program. We sync leads when they QL, and we sync leads when they're new.

And then think about your reporting and data needs before. And if it's too late, you didn't think about it before, think about it now. But it's really important to think about what, kind of, analytics your teams are going to want 'cause that might impact how you set up your Revenue Cycle Model or how you set up tracking for it. If you're wanting to do reporting on stage changes, there's a lot of different ways you can do it depending if you have a data science team or a developer that can help. But you can pull revenue stage changes out of the activity log. This is a snapshot that shows you the, kind of, details that are there. You can get the previous stage, the new stage, the date time, all kinds of information out of the activity log. You can use a JSON field where you can concatenate all of the different data points that are important for you when a stage changes and put it in this nice little JSON formatting. And then a DI team can export that and change that into a table, and you can use that for reporting. You can use custom objects to track revenue stage changes. You can use a Webhook to create a custom object. There's all of different kinds of ways you can do it. Date stamp fields is a really common one. You just want to be careful because the revenue cycle is not a flat linear journey. Ideally, hopefully, it's circular. You got people turning back. You got people going forward. So date stamp fields, you have to decide. Do you want to update the date stamp the first time they hit a stage? Do you want to update the date stamp every time? Is it reflecting the most recent? So you just have to think through those things. And date stamp fields can be really great. Even if you're reporting in other ways, they can still give you great visibility, and it's easy way to run some data queries. In Marketo, if you have Revenue Cycle Analytics, Revenue Explorer Reports, give some good visibility into how long somebody's been in a stage, what the inflow, outflow is, things like that. Corey showed the Success Path Analyzer that can show you some good information. Smart Lists are so key. You can do so much with Smart Lists. It's more than just how you target and segment your database and put in Smart Campaigns. You can do a lot of reporting with Smart Lists too. And of course, there's people reports, which I'll show you a snapshot of later.

For us, one of the things that was important is we wanted to know how many times people were turned back, and it's really hard to get to that if you just have a turn back campaign, and somebody asked you six months later, who else gone through that flow twice, right? So we implemented a count field. Every time somebody gets turned back, we count. And we do that for rejects and QLs and all kinds of things. So if you're thinking through this stuff ahead, it really makes it easier later when somebody comes to you and they say, "Hey, I want to know how many times people are getting turned back." It really helps. You can do a history field to show, like, every revenue stage change, and you can concatenate to the end, like, their whole journey. There's just so many ways you can do it. So just something to think through.

And then, when you first implement a Revenue Cycle Modeler, it is not a field in Marketo. It's not a field in Salesforce. You can query it, and you can see it on a person record, but you can't see it in the Smart List report. So you have to create a second field. We create it in Salesforce, so it gives visibility to Salesforce, and it gives visibility to Marketo. For us, we call it Lead Qualification Stage. For you, it might be Revenue Stage. But you do want to create a field and then write the Revenue Stage to that field, so it gives you extra visibility to what stage somebody is in, in both CRM and in Marketo.

Okay, now the fun parts. So we're going to go through some gotchas. These are things that might be wrong in your database that you don't know about. If you see one and it applies to you, raise your hand. Corey would probably toss you a ball or try to toss you a ball. So we'll see how it goes.

All right, so merges. Probably 80% of these slides are merges, and I'm probably going to talk fast 'cause we have a lot of gotchas. So Person Score and only Person Score is added together when a record is merged. So if you have record A with 400 points, record B with 400 points, and you merge them, guess what? Your remaining record has 800 points. For some people, that's okay. For my company, we don't want that inflation. So a pro tip, if you don't want that inflation either, you can create a Smart Campaign to reset the Demographic Score to zero, re-run demographic scoring, and then re-calibrate the Person Score so you get rid of that inflation.

Program membership can actually re-trigger scoring during a merge. If you have record A who's part of programs, and record B, who's not, and you merge A into B, guess what? It's going to look like record B is just now attending and registering and all of those things for those events. And guess what? That can re-trigger scoring.

We have a workaround for this. And funny, I was trying to find a snapshot for the slide, and in our own instance, we didn't have the little filter in place to prevent that. So fix that really quickly.

- I get a stress ball for that. - You want to do that.

But it's really easy. To prevent that in any scoring Smart Campaign that's based on program membership status, you need to add a filter that says reason not constraint that says reason doesn't contain merge. If you add that, when you merge two records, they're not going to re-trigger program scoring.

Similarly, in your Revenue Cycle Model, if you have triggers that are off score so like for us, a QL is 800 points. We'd love to talk more about that, but we probably don't have time. See me after. But 800 points is a QL. Well, we don't want person A and person B to merge and make a QL. That's not good for us. Even though we fixed the inflation, it still happened. It could trigger a QL stage. So also add to, your revenue cycle Smart Campaigns that are based on score that the reason does not contain merge lead database. And the reason we get a little more specific here and add the lead database to it is because sometimes anonymous records get merged in, and there could be some score there. And that's okay for us. But this is really important on your RCM. You don't want to have people QL-ing due to merges or moving stages due to merges. At least we don't.

Also, be thinking about the winning record if you're merging records, if it's in Salesforce or Marketo. When you have two records in two different revenue stages, the winning record is the revenue stage that stays. So if you have a record that's a QL and a record that is new, if the new record is the winning record, you've lost your QL. You've lost it in Marketo. Depending on your sales process and how they work QLs, they might no longer know that they have a QL in Salesforce. It can really wreak havoc. So make sure when you're merging, you're keeping account winning records. And in Marketo, if you're merging, the winning record is the first record you click. So I know how many people have merged in Marketo.

So you know when you merge in Marketo and you see and there's, like, one that's, kind of, highlighted, and it, kind of, looks like that's the winning record. It's not necessarily the winning record. So make sure the one you click first is the one with the stage that you want to keep.

So back to merging 'cause merging is one of the most difficult things to actually solve for at scale. So had a client that had over 200,000, duplicates and we were in a situation where just clicking a button and manually merging wasn't going to get it done. Mix that in with about 900 different fields that I also need to account for. So how do you solve for that, right? Well, if you guys know, I like to tinker with the API. So I built a tool that, kind of, allowed you to do that. But right here in front of you is the general concept of I can stack as many IDs as I want in the lead IDs and merge it directly into that ID. But it gets a little bit more fun than that because what I'm doing now is I'm actually jumping into data telemetry. So it's a three-step process really. So one is I can't merge data and have data loss. So how do you account for that, right? Well, the first thing you do is you go into that Smart List, and you pull all of your duplicate data out, first and foremost, and you include every single field that you think is going to matter, right? This is going to be your data set that is going to tell you who your child is and who your parent is, and then all of the other fields that are a mix between parent and child. So I can have a child that has data, and I can have a parent that doesn't have that data, but I want this one to have it, right? So you can create data telemetry with Pandas and with Anaconda and it basically allows for you to understand more about your data, take it out of the system and have more control. That is critical at scale because merging and duplicates definitely can be a very, very difficult task, right? So you need to think, how do I solve this if I don't have access to one of the many tools that are out there that can handle this, right? With $10,000 here, $3,000 here. Fun fact, this is actually free to run. It doesn't cost me any money to do it. So also, a really, really good way to save money for a company. But once I have all of that data and I have all of my IDs all ready to go, I can run them through my program, they'll merge all of that data and then the aftermath is after I have my parent record merged, I can then up cert all of that extra data that I had within my CSV that I created, and then I can create a bolstered parent record that has all of the information and has all of the qualifications that I need in order to maintain and avoid as much data loss as I can. Is it a perfect science? I would say it's probably not a perfect science, so merging is messy, right? So what does that mean? Get ahead of your duplicates at all costs. Try to avoid creating duplicates. They are very difficult to deal with and they create an issue where Marketo has to work twice as hard to qualify people because customer A can have a lead score, customer B, which is the same person will also have a lead score. Why is Marketo only scoring this person sometimes and then this person sometimes? Now instead of getting 60 points to an MQL, I have to do 120 points to an MQL, right? Because the system has to work twice as hard in order to qualify those people. Another instance that you'll run into is that there's an instance where a child record may contain compliance opt-in data and the parent may not contain compliance opt-in data. Marketo, I still have never been able to figure out why Marketo targets this one person and then targets this person, right? But if I don't have the ability to say, "This is the person I'm emailing but I don't show compliance opt-in data on them and I get audited, how do I solve for that?" Right, so you just gotto be really, you gotto think very clearly about why are duplicates in the system? A lot of times, Salesforce is obviously creating duplicates, and so you need to make sure that sales understands that when they create duplicates 'cause they're lazy or the business doesn't understand it, it actually creates a huge issue for marketing. Especially, when it comes to GDPR and especially, when it comes to compliance. So just a few little extra notes on data duplication. - Back to you, Jenny. - All right. I think we only have a few more merge slides.

If you use custom objects, and I love it that I was in the custom objects session, and YouTube mentions this. If you have custom objects, if you have record A has a custom object, record B has a custom object, merge them together. The record that didn't win, the custom object gets orphaned. So if you have custom objects, have a plan to re-associate those custom objects to the right record.

So pay attention. If you have massive merge efforts or if you're just running like RingLead or Demand Tools or whatever it may be on a weekly basis, watch for data loss, data value changes, accidental triggers. Lots of things can accidentally get triggered when you're merging records in Marketo. So be watching for that. This is a Smart List you can run. There's no filter in Marketo to say, "Hey, show me who was merged." If there's anybody from product in here, that would be great. But this Smart List is the closest thing you can get to it. Since Person Score gets added together during a merge, you can run this Smart List in your database, see how many people have merged. You can only see back 90 days. But check it out. Look at some records. See if you have any of these merge gotchas. So we have a few more, but who's seen one so far? Because we have some balls to get rid of.

No? All right. You guys are good. All right. Okay. - So-- - You just want a ball.

Yeah. All right. So when you're decreasing the score, be very careful. We'll go through this one quickly. I'll just walk through the one on the right. Let's say you are trying to decrease the Behavior Score by five points. Every week, you don't have activity. This Smart List on the right or this flow on the right shows you how to do that without going below zero. You just want to check first, is the Behavior Score less than five? If it is, set it to zero, don't subtract five points. Otherwise, the default is, hey, subtract five points. To make that even more efficient in your Smart List, you should not be reducing the score if their Behavior Score is already zero. So just some things to think about as you're decreasing score if you do that.

Probably most people already know this one. All over there. Well, who is it? Okay. Probably you guys already know this one, but when you're demographic scoring, you have to use data value changes and person has created triggers. If somebody comes in with a data value, then it's not going to trigger a data value change. So make sure you're doing this if you don't want to miss out on demographic scoring.

If you're turning back or rejecting a QL, this is one--Gosh, I think I caught in our own instance maybe three years ago.

Everybody's probably dealt with race conditions in Marketo, but when you're turning somebody back, you have to make sure you're doing it in the right order. For us, when we're turning somebody back, we're resetting their Behavior Score to zero. We're recalibrating their Person Score to equal the Demographic Score. And what I found was somebody had set it up to change the revenue stage first. Well, guess what happened? They got set back to engage, but they still had that high score for just a second or a millisecond, and they started moving back through the RCM. So make sure when you're turning back, you're taking that into account. So for us, we reset the score first. We do everything else first. We reset the stage last.

Corey already mentioned SLA stages are not business days. We do not use them because our own sales team does not like it when SLA starts on Friday and expires on Saturday. So if you're using SLA stages in your RCM, make sure everyone understands that how the expiration works. And then lastly, be careful with the running through flow every time, running through flow once an hour. If you have a Smart campaign with multiple triggers, you might want to figure out if maybe you should set it to once an hour, not every time. In our QL Smart Campaign, I think we have five triggers. We're looking for Person Score, Behavior Score. We're making sure you have a phone number because sales need to be able to call you. There's a couple other things. So we have five triggers and five filters. And most of the time, you probably would run through it once. But let's say you fill out accelerator form, you might be triggering all 5 of these things within 10 seconds of each other. And next thing you've got a record going through the flow five times. You don't want to send sales five alerts for one QL. So you can only QL once an hour. So think about those things on any Smart Campaigns in your Revenue Cycle Model that maybe have more than one trigger, and some other things in that area to watch out for too.

You could have another example is web activity. When it comes to bad QLs, I had a client once. He wanted to score five points for every web page visit. And before they went live, we had to talk to them and say, "Hey, you have people who maybe surf the website a lot one day." You don't want them to get a Behavior Score of 500 points for surfing the website for an hour. So a way to combat that is you run that scoring once every hour, and then you get a little bit better depiction of what that looks like. So just watch out for your qualification rules and think through what's the right one to use for you.

Okay, so we'll go into testing really fast. My background, my first five years is out of college, I think I was coding and QL-ing. So testing is near and dear to me because 13 years ago or 14 years ago when I came to the world of Marketo, marketers didn't really test back then. So make sure you're testing. Make sure you're testing to make things work as expected. Make sure you're testing to make sure other things aren't happening because of what you set up. And here's just an example of a testing plan, the way I document testing, and have my team document testing so that we know everything is going to expected prior to launch.

And we all testing production probably, most of us with Marketo. Make sure you're using test filters. We use naming conventions. For us, it's @test.com If you add this to your Smart Campaigns, you can testing production without worrying about what you're going to do to the rest of your database.

And it's not just testing what you've built, you also need to test before you've built. There's a lot of strategy that goes behind lead scoring and Revenue Cycle Model, you should be testing your strategy before you ever build 'cause if there's a strategy problem you really want to know before it's too late. So if you have a scoring model and you know that, hey, these demographics or this scoreand you know that, hey, if they fill a form, it's this score, you can already figure out, okay, they come to the website, they fill out a form. The demographics will give them this, and the behavior will give them this, and their score is this. You can already see before you build if the model's actually going to do what you want it to do. So you should actually be stress testing your strategy before you ever built. And of course, after, always, always monitor and watch for issues.

For launch, just make sure when you're launching something big, if you're building a new Revenue Cycle Model or new scoring model, newly management foundation, always have a launch plan. It seems really easy to just think, "I'm going to turn it all on, and it's going to be fine." But if you create a launch plan and you think it through, it's really going to help everything go smoothly, and it's going to help you catch those things that you may have missed. Order of events is really key, and thinking through things like if you're redoing scoring or if you're setting up a new score and you need to batch people through the demographic Smart Campaigns, if you have a large database, it takes time to run it through all of those Smart Campaigns. You might not want to do that on launch day, right? You might want to do that the weekend before, or the day before, or night before. So just think through those things and always have a launch plan.

And then, for us, this is an example of one of the things we documented for ourself, but I took Corey's Revenue Cycle Model in there. But you always also have to think about starting stages. When you have a new RCM or you're changing the stages in your current RCM, you gotto think about where you want people to go. Marketo doesn't automatically put people in it when you start or activate an RCM. You have to put them in the starting stage. So make sure you have definitions defined and that all of the teams agree, like, who should go in the opportunity stage, who should be in MQL to start, who needs to be at the beginning. It's really important to just document and have team alignment on that.

Okay, optimization. So we've talked about these things that can wreak havoc to your RCM. So let's talk about how you can watch for those.

When you have a live RCM, make sure it's working as expected. You can use Smart Lists. I've got snapshots of some of the Smart Lists we use on the left. We use Smart Lists to monitor all kinds of things, making sure people are in the right stages. The Success Path Analyzer is at the top that can help you find bottlenecks. And this little report snapshot is from a sandbox, so the numbers look a little weird. But this is a great report to run. It's a person status by report. You can use custom columns across the top to show the Marketo Revenue Stage somebody's in. And on the left, for us, it's called Lead Qualification Stage. That's that custom field. And then you can see if anybody's out of alignment. Their Revenue Stage and the custom field you created should be the same. But over the years, I've seen, "Salesforce didn't lock down the field and a salesperson changed it." Or, "A record merged, and somebody selected the wrong things and changed it." So that helps you find bugs that way. It can help you find all kinds of things.

And so here are some things just to consider and watch out for. Watch for people getting stuck near thresholds. So I mentioned earlier we have 800 as our QL threshold. But a little while back, I found that a bunch of people were at 790, 785, and it was preventing people from QL-ing. So I went to our strategy team, and I said, "Hey, I think we've got a problem here." Sure enough, we looked at the scoring model, and, yep, we needed to increase the score a little bit. Those people should have been QLs. So watch for your thresholds. If you have every stage along your Revenue Cycle Modeler, look for people that almost meet your criteria. And if you see a lot, maybe there's an issue there.

Looking for too many turnbacks or rejects from sales. Maybe you need to change your QL definitions. Look at too many people getting in Lost Opportunity. Maybe you want to turn them back earlier in the model. Maybe you want to nurture them, depending on what's good for you. Look at people who are lacking in progression. If you don't have people moving through the funnel, maybe there's something wrong. And then for scoring, look at people who have negative scores. Look at if you have somebody with a score of 10,000. I mean, that could be interesting. Look at that.

And then if you have a person that has all of the demographics, but their Demographic Score is really low, maybe there's a demographic issue. So there's all kinds of different ways you can look at it. You can also look at somebody's filling out forms, but they're still in the known stage. For the way, my model works, if you fill out a form, you move to the next stage. If you fill out another form, you move to the next stage, right? I think four or three form fills, you might be a QL. So if you people are filling out forms, but they're not moving stages, maybe you want to look into that. There's an easy Smart Lists you can use. You can just say filled out two forms, and they're in the first stage, and then look at them, see what's wrong. There's lots of different ways you can dig into it.

Okay, so some key takeaways.

Strong strategy is crucial. Make sure you incorporate the full, circular journey. Make sure the Revenue Cycle Model is more than the path to MQL. It's important to track all the way through opportunity and make sure you're capturing those circles. Follow best practices to make sure your lead scoring and Revenue Cycle Model aren't prohibiting you from driving growth. Watch out for all of those gotchas we just covered. Merges, especially, can be tough. And always watch and monitor and always continue collaboration. Revenue Cycle Model and lead scoring isn't set it and forget it. How many people-- And it's okay to raise your hand for this. How many people have a Revenue Cycle Model and lead scoring they haven't looked at in, like, over a year? Yeah. Yeah. Maybe go home and take a look at it. Do I throw a ball to those people? Yeah. All right.

So still have questions after today. There is a Champion Office Hours next month with an amazing panel. I'll be moderating. Hey, if you're on the panel, stand up. I know-- Yeah. So come to the Office Hours. They'll be happy to answer questions. Also, Corey and I are in LinkedIn. We're more than happy. We love helping the community. So if you have a question or anything, we're here for you. And I think it's time for Q&A. So thank you so much for listening. And if you ask a question, we'll throw you a ball.

[Music]