How to Model and Predict ROI from Content Marketing by David Khim (Omniscient Digital)
Bernard Huang
David Khim, cofounder of Omniscient Digital, led a webinar on how to predict ROI for content marketing.
Here are our biggest takeaways from David’s talk:
Models give you an idea of high-level growth opportunities
Err on the conservative side with your estimations
Do not make your models too complex
Models are mostly assumptions
Watch the full webinar
Check out David’s slides here.
And check out the resources David shared below:
About David Khim:
David is co-founder of Omniscient Digital and Head of Growth at People.ai. He previously served as head of growth at Fishtown Analytics and growth product manager at HubSpot where he worked on new user acquisition initiatives on the marketing and product teams.
Follow David on Twitter: https://twitter.com/davidlykhim
About Omniscient Digital:
Omniscient Digital is a premium content marketing agency that helps B2B software companies turn content into a growth channel.
Follow Omniscient Digital on LinkedIn: https://www.linkedin.com/company/omniscient-digital/
Read the transcript
David:
All right, so hey everyone. Thanks for making it today. Today we're going to be talking about how to model and predict ROI, return on investment from content marketing. One of the guiding principles for my business, Omniscient Digital, is to level up the content marketing universe. And the reason why we got to this principle is I think a lot of the conversations we typically have in the content marketing space tends to be around, "Hey, is this high quality content? Is it long enough? Are we using Oxford comma?" And while a lot of those are important, I'm seeing a chat come through here. Let me just see what's happening. While a lot of those things are important, it's missing the picture of the marketing side of the content, right? Is this content going to contribute to business goals? And this is something that I had to learn firsthand during my six years at HubSpot and growth, and it's something that I hope to teach folks as well.
This has been one of the most valuable skills for my career and hopefully it's helpful for you all as well. So really weird formatting, but as Travis mentioned, I'm the co-founder of Omniscient Digital, previously head of growth at People AI and fishTIME Analytics. Spent six years on growth at HubSpot as well. Some of those logos at the bottom are companies that we've worked with to help them grow traffic leads and revenue through content and SEO. So today we're going to go through the basics of building models. I will note that this is not going to be a 101 session. If you have not built a model before, then you might be jumping straight into the deep end. But if you have any questions, we'll do our best to get to them. I'm going to go over some mistakes when building models that I've seen over and over again.
We'll also have a couple things that you should have established before you even touch a spreadsheet. And then we'll do a walkthrough of the model. I'd like to do a live demo. So maybe for some folks here, if you are open to being in the spotlight a little bit and just having your content strategy or business in a limelight, drop your URL into the chat and we can potentially use you as an example for the live demo.
All right, so I want to start with a story here. It started off with this concept of, I just had an idea. During my time at HubSpot, we were part of a spinoff within a company called Sidekick. It turned into HubSpot Sales. We got merge into the made marketing team. And candidly, they had no idea what to do with our team. They were like, "We don't know what you were doing. We have our team here and now you're just a bunch of five folks we have to figure out how to fit into our team." I didn't really have a role to find anymore.
And after maybe a week or two of that, I just said to my VP, Kieren, I was like, "Hey Kieren, I'm going to pitch you what I think I should work on. And I know you don't have a clear path for me right now, but I don't like just sitting idly. So let me just pitch you. Give me two weeks to do some research. I'll get back to you". And then I came back with my research and I forgot what the exact ideas were. I think it was something like, I think we need to work with product on integrations more. I think we need to build some more courses separate from HubSpot Academy.
I think we need to do a content strategy that supplements what the blog team is doing, but fills a gap, which is what we're going to get to here. And I think the fourth one was some virality through our free products that the team wasn't doing yet. So had a bunch of ideas and surprisingly Kieren was like, "Yep, go for it. Run with it. Let me see what happens." And I was like, "Oh great. Good news because I've already started." The thing that happened after that was he said, "Hey, wait a minute actually before you run off with all of that, what's the ROI we should expect from all of those things?" And I was a bit earlier on in my career and I was kind of a dear caught in the headlights. These were a lot of things that the company hadn't really done before.
I've seen other companies do them. That's how I got the ideas. But we didn't really have a baseline internally for what sort of return we could expect. And I'll use the content strategy one because that's what we're talking about here. It was really hard to say, "Hey, what's the ROI of this content that might generate more leads? What's the ROI of content to help people evaluate CRM and marketing automation tools?" I don't know. The blog team was riding very top of funnel educational stuff and they're focused on traffic. In my pitch I was saying, I think we can get more leads in product signups. And so he was like, "This is very different from what we typically do. If it's not traffic or keywords you're talking about how are you saying we can get leads from this content? What's your baseline?" And so this was my first foray into learning how to build models and being able to predict like, "Hey, what can we actually expect from this content that we're doing?"
And so let's get into some of the basics of building models. Just some things to keep in mind as we go into this. So this is a model. This is a model and this is a model. You'll notice it's just a bunch of spreadsheets and numbers. These are actual models I've built. This is when I was trying to decide if I should buy an investment property or dump that money into index funds. I bought an investment property but I had to do some numbers to feel good about it. Probably wasn't best, but now that we're seeing what's happening with the stock market, I actually feel pretty good about this decision. This was me building out a model to understand what my stock options would be if the company I was previously at IPO'd. These are fake numbers and I had made some really big assumptions on what the stock value might be, but it helped me have a better understanding of how many of my options I should exercise and what I was comfortable with.
And this is a model example from reforge.com where you're trying to model out the impact of creating a viral invite loop. And you'll see that there's a lot of numbers, a lot of them are assumptions, some of them are based on reality, some of them are just kind of ballpark estimates, but that's how models work. There are a lot of assumptions. And this is where we get to the first big point I want everyone to take is the goal of building a model is not to build out what reality might look like, it's to just have an idea of growth opportunity. You are modeling out but one of many infinite scenarios for what this investment might look like.
I like to quote George Box here, one of the greatest statisticians of all time. "All models are wrong but some are useful." So let's keep that in mind as we go through this. Some of these are mistakes I've seen over and over when building models. The first one is being very generous with estimations. I've been part of conversations where I'm like, "Hey, currently our baseline conversion rate is about 2%." And with that conversion rate, someone might say, someone in leadership might say, "Oh that looks a little low, can we make that 2.5 or 3?" And I'm like, "That's not reality though. You're playing with numbers in a spreadsheet and you're creating some future that you want to pitch, but you're going to set yourself up for failure. So I always be lean on the conservative side of my estimations and some people might say that's sandbagging. I say I'm building in buffer for things to go wrong because things always go wrong.
Folks also end up spending too much time building models as well. They're trying to think about all the edge cases, all the other assumptions they can layer in and their results in spending too much time and building a very complex model to the point where if you're trying to explain your assumptions to someone, they're going to get very confused and very caught up in a lot of the details.
So my MO is keep it simple, don't get fancy. If you build a simple model and you start running with it and setting goals against that and later you decide you might need some more complexity, then great. But starting with complexity I've found is always a way to plan for things to go wrong. And so this sort of sums it up really is don't spend too much time scrutinizing specific numbers to try to build a perfect model. There's never a perfect model.
What I like to do is give myself a week to gather inputs and information, chat with a bunch of people. In this case if it's about content like, "Hey, what are you seeing overall website conversion rate looks like? What are you seeing to blog conversion rate looks like? Are we actually working on improving a conversion rate and what's our traffic average growth been to date?" Just to get some of those numbers and I'll give myself maybe another week to build it and say, "Hey, we don't need to spend more time on this. It looks like there's going to be a good amount of opportunity here. Let's move forward with it." Are there any questions I'm missing here before I go further?
Travis:
No, we're good.
David:
Okay. I'm just seeing a red chat button but it's not popping up for me. So now we're going to get more into the weeds of some things you should have established before you build a model. So first, what's your goal of investing in content? Lots of folks have trouble communicating this. It might be we want to grow organic traffic, but is that really the goal? I'd imagine if you're doing content for a software company, the goal is probably to get leads and eventually revenue. So being super clear about that. And your goal could be to become like the go-to educational resource where you might not care about converting traffic immediately, but having that explicitly stated from the start is going to help avoid a lot of headache down a line.
I've heard people say, "Yep, we want to be the educational resource for our industry." But a couple months later they're getting asked about leads generated from your content, and sometimes those goals can be conflicting. And if you can actually say, "Nope, our goal was not generate leads, it's to focus on top front of traffic and building a moat." That's a very valid goal and it changes the way you build a model.
The second point is you should at least have a general idea of what your constant strategy may look like. You don't want to go start building a model without actually knowing what your strategy's going to be. The model comes later when you've decided, okay, we have an idea of what our strategy is, let's see if how this plays out once we start adding numbers to it. And with that, have some high level topics of what the company strives to be associated with. That might be high level broad phrases that your company wants to be known for that's related to the product. It might be philosophies you want to be known for. If you want to use HubSpot as example, they want us to be known for inbound marketing. And then, more product related, they want it to be known for CRM. These are really the high level topics that you should have an idea of that allows you to do some broad keyworded research into those topics.
And you don't need to do very comprehensive research to start using the model, but it helps to have some of those keywords and search volumes in place so that you have some numbers to work with. And you'll see that some of the examples I use are going to be placeholders. I'm going to copy out right now. I do not recommend going after these keywords, but they're just going to give us something to play with.
The next thing that you go on is some general conversion rate ballpark. So what's your website visitor to lead conversion rate and what's your lead to customer conversion rate? If you don't have exact numbers, you can use something conservative until you get a hold of right person to get closer number is to reality. But, in the meantime, those are the minimum two numbers you should have. And then the ballpark average contract value or customer lifetime value. I prefer average contract value. It just that if they sign a contract this is what they're paying you, whereas customer lifetime value assumes that they're going to renew and all these things over many years. And I just feel much better making assumptions about the short term rather than the long term. Some folks have a different approach. We will be using average contract value in our model today.
All right, so I'm going to switch my tabs here. I'm going to do a quick check on the chat. Sweet. All right, it was just Travis and Bernard bullshitting. Just kidding. All right, so let's get into this tab here. So you'll see here that there are really four tabs. I added this fifth one for example, that will be pulling. But once you get this sheet, you're going to see this tab and it's just going to tell you only edit the yellow cells. You can edit everything else. It might break the sheet, and just don't request this because I'm not going to pay attention to those requests.
So first we have this sheet, it's very straightforward. It's going to be just whatever keywords and search volume you have for those keywords, paste them in here, and it's going to populate the rest of the model here.
Travis:
I do see a question from Josh. What if your ACV file varies wildly by product line? Ours can be anywhere from 3k coming up to 550k per month.
David:
That is great. What I would recommend you do is you might want to have a couple different scenarios where you might have a content strategy that maps to one product line, another one that maps to another line. If you try to combine those, which I've done for two different business models before, it gets very complex. So I'd I'd recommend avoiding doing that. But you might just have to have a couple different scenarios mapped out in your models.
All right, so I'm going to walk through this before we populate this sheet. Actually no, let me just put these in. So I have some example keywords from a website here. Let's just assume that these are good keywords we want to go after and that there's some business value to them. You'll see that a lot of these I pulled. What website did I pull these from? I think I pulled these from HubSpot actually. So again, I'm not going to say that any of these are good business keywords to go after. They're just going to give us some data to work with.
But once you put your keywords into this sheet, they get pulled here. So this is the traffic assumptions sheet where we have assumptions for clickthrough rate based on your ranking on Google. So I'm sure these numbers may vary week to week or even month to month. But if you're ranking position one, what this model says is you're probably going to capture about 43% of the search volume for that keyword. If you're ranking position two, you're going to get 14% of the clicks. Position two, you get 8%, and position four you get 6%. And so here, let's say you rank for the keyword business strategy and position one, the search volume is 5,000 per month. You might get 43% of that, which is going to be about 2000 organic visits per month from that keyword.
If you're in position two, you might get 700, position three, 400, position four, 300. So it does that for every single keyword here. And again, these are assumptions. They may not be reality, but at least gives us a ballpark to work with. When we look at this little table here, what we're seeing is let's assume you rank position one for all these keywords. This is the total amount of monthly search traffic you'll get from these keywords, which is a lot. And assuming you can rank in position for one for all these keywords, it's also a very aggressive assumption and we'll get to that later. But we do the same here for position two, three, and four. That is this sheet. This is just to set the baseline of how much traffic can you expect.
All this now gets populated into this growth model. So I'm going to zoom out a little bit because there's a lot happening here. So up here we're setting some variables where we're saying, "Hey, currently we get about 10,000 organic visitors per month to the website. Let's just use that. And let's say that the visitor to lead conversion rate is about 1% and this adds a little layer of complexity, but let's say that each month we improve the conversion rate by 20 basis points, which is .2%. That's a pretty good optimization if I had to say. So what I'm going to do is I'm going to put that to zero. Let's assume you're not improving your conversion rate, you'll see that the leads change immediately. But let me walk through how the rest of this sheet works. So here in position one, what we do here is we assume linear traffic growth.
Why? Because it's easy. When you start investing in content SEO, of course the goal is exponential traffic growth. Can't really predict that. And so I'm going to save us all the time and say let's aim for linear growth because that's going to be easy to map out. And so what we do here is, let's say this is over 14 months. Let's say within 14 months, what is it? We're going to add 107,000 organic traffic visits per month. So we're saying if we're at 10,000, what does linear growth look like to get to that? That is it. There's no fancy calculations here. We just take that number, divide it by 14 and add up across the months. So then that's how we get this blue line here of traffic growth for position one rankings. We do this for position two rankings, position three, and position four.
So now what we can end up seeing is, okay, well going from 10,000 to 117,000 organic visits per month sounds way too aggressive. Maybe that's not in the realm of possibility, but maybe aiming for position two or even three is more realistic. So now if you're communicating to leadership, they might be saying let's 10x organic traffic. If you look at this line and you look at historical growth, I'm willing to bet that historical growth is like flat. So now if you can push back on leadership and say, "What leads you to believe that we can 10 x this at our current investment," which we'll get into in the next tab. But the next layer that we'll add here is we have this 1% conversion rate pulling from this cell C10. And you can see that if you get 10,000 organic visits a month at 1% conversion rate, you get a hundred leads per month.
And that's this blue line that's mapping linearly. So now you can see, okay, if we have a baseline conversion rate, we can kind of predict how many leads we're going to get from that organic traffic. And you can see that if we increase this by, let's see, when you basis points, that's going to be exponential growth. So even though your traffic growth stays the same because your conversionary gets better month over month, you can expect to get more leads from that same traffic. This doesn't have the nuance of well is your content top of funnel or bottom of the funnel? That's a whole different conversation. And the only thing I'll add to that is just because your traffic grows does not necessarily mean your conversion rate goes down. I'll say that again, just because traffic goes up doesn't mean conversion rate needs to go down.
The reason that happens for a lot of folks is because the content they're writing targets very top of funnel 101 content where it's very educational and helpful, but the people reading that stuff are never going to sign up for your product. If you go to our website beomniscient.com, we have a case study with Jasper that shows this where we grew their traffic by crazy and their conversion rate from blog to product sign up went from 1% to 2%, 3% to 6%. And we were like, "Maybe this is a fluke, maybe it'll go back down." But we had the hypothesis that if we're producing bottom of the funnel content that targets people with high intent to convert, we're going to be able to maintain that 6%. And we have.
So that's an assumption I want to challenge folks here who may be used to seeing low conversions from the blog, that doesn't have to be happen to be the case. It just so happens that people are creating a lot of top of funnel content that is not going to convert and that's why they have low conversion rates.
Travis:
Yeah, Dan has a question kind of on that piece about the, how do you account for session to lead conversion rates for keywords representing different stages of the funnel? How would you approach that? Are you saying only focus this kind of method on bottom of funnel content?
David:
Yeah, I think there's two approaches. You could make this model a bit more complex and you might have a list of keywords that are more top of funnel, a list of keywords that are middle, a list of keywords that are bottom and have a different conversion rate for each of those clusters of keywords, and then model your growth that way. Or you can say, "Hey, for this next quarter we're going to focus on bottom of the funnel content. We're going to publish 20 articles a month. We see that bottom of the funnel content converts at 5% and top of funnel converts at 0.5%." And then you can model out, okay, this model is only for our bottom of the funnel content, it's going to be lower traffic growth, but the leads are going to convert, or the traffic is going to convert at a higher rate.
So it's really a decision, Dan, if you want to put all that into one model or if you want to have a separate model for each one, I recommend keeping them separate. But I can see a case where you or your leadership team might say, put this all into one thing, which is a whole different exercise. But hope that's helpful. It's really dependent on what you're comfortable with in your approach. At least that's what comes to mind right now. I might have a better answer in a week.
All right, so this is the growth model for traffic and leads. So I think the key thing now is, well, where does the ROI come in? So that's where we have this break even analysis tab. So again, this is the lot, we'll walk through it piece by piece. So the first one is let's say you have a monthly content investment and let's say it's $5,000 a month. Maybe you're paying an agency $5,000 a month, maybe you have a salaried 60K content marketer that you're paying 5,000 a month, whatever it is. Maybe it's higher if you have a team and they have a percentage of time spent on the content program, but we'll keep this simple for this and just say you're spending 5,000. Let's say your lead to customer conversion rate is 1% and your average contract value is 1000.
Okay? So those are the constants that we're working with. Now for this next section, let's only look at position one. I'm going to collapse everything else for now. So this is pulling from this previous tap. So organic traffic, the conversion rate, the expected leads, this is all pulling from this traffic, this conversion rate, and these expected leads per month. All right? And then now this monthly costs, we're pulling from this cell, 5,000. The cost to date, we're just adding this per month to date. And the cost per lead is that cost to date divided. Actually, how did I do this one? It's in that month. So if you're spending 5,000 that month and you expect 200 leads, you're spending $23 a month. And then from those leads, you expect 1% of them to turn to customers. And at that ACV, so hundred leads, one customer at 1000, your predicted revenue is 1000.
And then we have organic revenue to date. So these two columns here, M and N, are probably going to be the ones you'd want to pay a lot of attention to. Because what you'll see is on the monthly return column, that first month, you're losing money. No surprise. You're spending $5,000 on all that content. You got one new customer, you only get a thousand dollars. You have a negative 80% return. And the next month that looks a lot better, and the third month looks a lot better. And in the fourth month, finally, within that month you've broken even. So in that month you made 5,500. You spent $5,000. But year to date, you have not broken even yet because over those four months you've spent 20,000 and yet you've only made 12,000. And so that's where here in month six, this is what you want to pay attention to.
So month six, you spent 30K on a content program. By month six, ideally you've gotten 10 customers. Actually no, you've gotten 30 customers and you've made 30K. So now you've broken even on your investment and the leads that you continue getting, you're likely going to be making more than you're spending on your program. Now this is assuming you're ranking position one for these keywords. So we saw how aggressive that growth was. When you look at position two, it takes seven months to break even within a month. And it looks like here it's almost a year to break even on your actual total spend. And you can imagine position three and position four look even worse. Yeah, position four after 14 months, you don't break even.
So let's play around with these numbers a little bit and say actually our average contract value, I think Dan, you're saying, let's see. You said 3K per month. So over a year, I'm going to assume it's an annual contract. It's 36K. If you're spending 5,000, you're getting one customer, that's going to be amazing. But let's say that, hey, these customers actually have a long buying cycle, or it's a small audience, so it's actually very difficult to get in front of them, convert them, or there's a big buying group that you need to convince. And let's say you need to spend... 10,000 seems small for that. Let's say it's 20,000. Even then it'll be good. But then this also leads to the question of, well, if your average contract value is 36K, they're probably staying for longer and you're probably willing to spend a lot more than that.
So this is getting to the question of depending on your business model, Dan, maybe you want to use lifetime value. Maybe your average contract length is like five years and you're willing to spend 100K to get one customer. So this is where we get more nuanced and really dependent on your business model. If we were to use a different, maybe like SMB software as an example, let's say someone's going to spend 5,000 a year and you can spend 10,000 a month on content, and actually your conversion rate is not that great and it's 0.5. That's not right.
Then you can see how that changes numbers a little bit. So hopefully this is starting to paint a picture of understanding what the ROI is on your content program. It's really dependent on a lot of these values. And you can get much more complex with this. I know some clients are saying, "Hey, we want to invest up front." So we have a client that's saying, "We want to go hard on content for six months. We're willing to invest 20 K over six months knowing that we're not going to get that payback period until later. But because we're frontloading all our content, we're going to be able to see faster results from SEO and traffic than if we were try to do two blog posts a month for two years."
It's very different to do 20 blog posts a month for three months than two posts a month for three years. Probably get about the same amount of content, but because your velocity is so much faster, the variables you're playing with and how quickly you get return is going to be different. So those are also things you want to consider. All right. Did anyone share a URL for a live website?
Travis:
Yes, we got one from Jean, getlasso.co.
David:
All right, thanks for volunteering, Jean. They have marketing tools for bloggers. All right, So don't take anything I do here as a directive of what you should do. I do not take any responsibility unless if you make money, then I will take all responsibility. So I could do some raw keyword research, but I think I'm going to do a shortcut and just say your competitor is Partner Stack. Oh, and I hope that's accurate, Jean.
Okay, so what I'm going to do is I'm going to cheat and just say, actually I know Partners Stack content is not great. I know this because I audited them recently. So Jean, if you are still here, what's a competitor that you frequently look at? Or if y'all know any other affiliate software, Thirsty Affiliates. Okay. How's their content look? Okay, they have some keywords. And what I'm taking the approach of right now here is let's just assume you steal their keywords to make it simple. So this gives us some stuff to work with. So you're all going to get to see me do this stuff live, so no pressure. So here I'm just going to import those keywords. So essentially, Jean, what I'm going to be doing here is let's assume they're going after the right keywords and you're thinking, should we try to steal their keywords? We're not.
It's a lot of keywords. I'm just going to get the first hundred here. And I'm going to delete those first key ones cause those don't look helpful. Okay, so we got those keywords and we're trying to decide who we just want to steal, Thirsty Affiliates, keywords, and if that's going to be reasonable for us. So you were saying, okay, if we stole those keywords and got position one, we could add a hundred thousand organics per month. What I'd recommend, Jean, is if this is a competitor you're looking at, see if any of these make sense for you. I can already see some you probably don't want to rank for clickbank.com. That doesn't make sense. So there's some cleaning up to do. But let's assume that these keywords all make sense. What you can now start playing around with is, let's say you're at zero organics, which I bet you're not at, but you can start seeing based on these keywords in a search volume, can this make sense?
And so here, let's assume that we're not improving conversion rate either, and your visitor to lead conversion rate is 1%. So you can probably start seeing the numbers play out here, but I'm just going to try to make some assumptions about your business model.
So I'm going to assume that your average deal size is probably $84 per month. Let's just say it's a hundred per month. Oh, let's say it's a thousand. So most people are paying you a thousand per year. So let's just put that here and assume it's a one year contract. And then let's say you're investing 10,000 a month. I'm not sure if any of this is correct, Jean. So if you're investing $10,000 a month and your average contract value is a thousand, even if you rank in position one for these keywords, you're not making much back at all. You don't really break even within a month until 13 months later.
So a couple things I'm going to try to understand is, well, can you do more with less of a marketing investment that makes your return on investment happen much quicker? Can you improve your conversion rates to 2%? Can you go after different keywords? Is your average contract value actually higher? Should you be improving that so that you can justify your content investment? Maybe it's actually, what's the other?
So this one, I'm going to assume some people are actually on this. So I'll even just increase this to 2000. So this might lead you to some different questions about your business model. Maybe it's worth increasing some prices and getting the ACV up so that you have a more evergreen organic machine coming in that can get people at a higher contract value. So these are of the questions that I would ask. Of course, it comes back to are these even the right keywords, and can our conversion rates actually be better than 1%?
That's something I'm going to continue harping on because if you're producing the right type of content that people are looking for when they're evaluating affiliate marketing software, you could reasonably get a higher conversion rate. If you can speak to the pains that people have using their Seed affiliates or PartnerStack and convince them that your product is the right one, I'm willing to bet you can have a better conversion rate than 1%.
All right, glad that was helpful, Jean. I think we have time for one more. One thing I do want to mention before is if you go to this link, this goes straight to booking on my calendar. I mentioned to Travis, I'm happy to spend time with y'all and help you work through these models. I had a lot of help when I was learning how to build these things and how to question my own assumptions. So if this is helpful, more than happy to spend some time with y'all.I'm not going to sell you on anything. I just want to help, because I think this is really important for us to do in the space. And in the future if you want to work with me, I'm up for that too. But I won't be doing a hard pitch or anything on this. I'm just here to help.
Travis:
Awesome. Yeah, thanks David. And as we're kind of wait for others to drop in their links, how do you apply this to this model to your work at Omniscient?
David:
Yeah, so Kathleen, Davey. So I am not actually going to do 360Learning because they're a client of ours. But, I will do a competitor if you have one in mind. I'm curious why Kathleen's asking. Kathleen, do you work at 360? Ah, Docebo, we're coming for you. All right, let's take a look at that. Kathleen, do you use AHS? Cause you can just pull all this data yourself anyway, but let me just take a look at y'all.
All right, so there's no pricing. So I can't really make an assumption on what your ACV is. Let's see, let me just do this then since you use AHS and you can just pull this sort of reporting anyway. Oh, LearnUpon, never heard of them. Google has so many products I don't know about. I just realized they have an Airtable competitor, which is fascinating. All right, so let's do learn upon here. Oh, they rank for some interesting keywords. What are these? I'm going to cut these down to something a lot more reasonable. So this one is a bit messier because they're ranking for a lot of keywords. So I'm just going to do some quick filtering.
All right, I think I saw you gave another example, but I'm deep in this one and I got to hop in 10 minutes. So we're going to run with this one. So here we're going to do an example for LearnUpon. Oh, it's weird, it's not loading. Want to export. Import.
Okay, there we go. All right. So these are keywords that LearnUpon is ranking for. I will always add this caveat. Do not assume that keywords competitors go for our keywords. You should go out there. You will find out lots of competitors also don't know what they're doing. So let's add these in. And Kathleen, I saw that Docebo is an enterprise software, so I'm going to assume that your average contract value is huge and that also means you have a big budget for content. Let me know if that's incorrect. I don't know what enterprise LMS has charged. Let's say it's 50K. And then let's say you want to invest 50K into content. And because it's enterprise, maybe your lead to customer conversion rates are lower. So if it was .3%. And in the growth model, let's say that you're currently getting about 10K visits per month.
Cause I think Docebo, y'all are pretty established, right? So let's see what this looks like in the break even analysis. All right, so assuming that you're spending 50K a month in content, because you probably have a team and they're spending some percentage of the time on a content program and your average contract value is 50,000, maybe it's higher. And if it's higher, great. And your lead to customer converging rate is, I'm saying .3% only because coming from enterprise software, you might have a lot of leads, but they may not be ready or it's a long deal cycle. So I'm just saying it's kind of low here.
So if these are reasonable assumptions, you may not be getting a break even month until month six, and you may not be breaking even until, what is this? Month 10. Again, I think if your average contract sizes are 100k, this changes what things look like a lot. And I think what's your domain rating? Cause that's going to change a lot too.
You have an 80 domain rating, so you could reasonably go for very competitive keywords and rank and position one or two pretty quickly. So that's great news for you. So I think it's reasonable to say, "Hey, we can aim for position one or two for a lot of these keywords. You're established, you have a high domain rating. Most content you write, you're going to show up on page one probably, assuming it's optimized using clear scope of course." And that it's high quality content. So this is pretty great. I don't know if you want to use LearnUpon as an example Kathleen, but that might be a company you want to do some keyword research on and competitive analysis on.
Travis:
Awesome.
David:
Awesome.
Travis:
Yeah, we got a question from Josh. He's talking about the model and base is asking, yes, we can see position one, two and three and four, but he is kind of asking, what's the blended approach? How do you model real world outcomes if you're ranking for one for a couple, two for another, and then three and four for some of the other keywords? How do you come approach that?
David:
Yeah, my first recommendation is don't worry about that. I'm willing to bet that you're not creating a constant program or SEO program and saying you want to rank in position 10. But if that is a scenario you want to think about, then I'd recommend in this tab change these estimated clickthrough rates to what you might see for, I don't know how updated this is, but what the clickthrough rate might be in position 9 or 10. It doesn't look like. So 10 might be a 1.5% clickthrough rate. Looks like these might have been updated recently. So I would change those assumptions.
If you're talking about blended averages, that's where it gets complicated. If you're saying, if these keyword rank in position two, these rank in position three, these in four, that adds layers of complexity. What I would recommend, and what I would do for myself, is just choose one that's like the average.
So let's say on average, yeah, we're going to rank on position two for some keywords or three, some keywords might be position 20, some are position 11. Let's say on average we're in position six or seven. That makes the calculation significantly easier versus, as I mentioned in the beginning, you don't want to spend a bunch of time thinking about all those different blended rankings. You want to have some simple model to give you an idea of what that opportunity is.
So I hope that's helpful. I know I didn't answer your question directly, but that's one of those things where I've gotten these questions from C levels and VPs and I'm like, "I hear where you're coming from. Very valid responses, not going to change the model very much. It might help us get a bit more accurate, but there's no perfect model, so what are we trying to optimize for?"
Travis:
Gotcha. That's helpful. And then Bernard has a question from the early part that I skipped over. How do you account for featured snippet and ads for CTR estimations?
David:
I have not accounted for those here. So that's a good question. Yeah, What do you think, Bernard? I mean, I think featured snippets, sometimes there's not even a click on those, right?
Bernard:
Yeah, I feel like overall organic clickthrough rates are so dependent on just the actual screen real estate that's given to the SERP. And I think there's maybe some creative ways that you could fetch that data through Ahrefs where organic search visibility, I think, is a percentage or a metric that they give. And I could see how that could factor into something that you might put into your model.
David:
Yeah. So do you think the presence of additional featured snippets might... This is a leading question. My assumption is it would decrease the click due rate.
Bernard:
Agreed.
David:
Yes. So I don't know what sort of buffer you would add or folks might want to add, but maybe 43% for position one isn't reasonable. Maybe it's more like 30. I think a couple years ago it was close to a 30, but seems like the... Oh, this is actually 26 now. This is changing every couple months. So keep an eye on this stuff if you're building out these models.
Travis:
And then we have another question from Andrea. Does the model calculate only search volumes or can traffic potential come into play? I think they're asking, we're seeing some low volume keywords bringing a much higher traffic than expected by the volume.
David:
Yeah, that goes into the whole low to no search volume keyword thing, which is, that can be a whole different conversation. So Andrea, I'd say I don't have a good answer to that. What I might even say is if Ahrefs is saying that it has 10 search monthly searches per month, but you're seeing a hundred visits per month, those are going to be things that you'll have to make a judgment call on. And it's not a good answer, I'll be honest. But it's one of those things where all those tools, I use keywords everywhere as well. And search volume on that is drastically inflated compared to a Ahrefs, so I use Ahrefs because it's conservative. But it'll say a lot of keywords don't have search volume when in reality they do. So I'm not sure what your situation looks like, Andrea, but that might be something where you might assume, if Ahrefs says there's no search volume, you might just say, "Okay, I know Ahrefs says this, I'm going to assume it's a hundred searches per month." And build that into your model.
Travis:
Awesome. Super helpful.
David:
I've had a similar situation where earlier on in my career I said to a client like, "Oh, I see you're raking for this keyword, but it has zero searches." And he said, "Actually, we do get traffic and it's our number one lead gen page." And I was like, "Oh, I sound like an idiot now." So I don't make any assumptions on this things now.
Travis:
Awesome. But I have to run a couple minutes. Definitely thank you for your time and we'll drop in the model and chat if you want to make a copy of that. And we'll send out the recording, plus the model, plus the slides tomorrow and then also a link to David's calendar in case you wanted to grab some time with him to kind of walk through the model. But David, anything else you wanted to share before we give everybody time back of their day?
David:
No, I think I really appreciate the questions. They get very nuanced and the only reminder I would have is don't spend too much time playing around with numbers and spreadsheets. Build out your content program, track how it's performing, and use historicals to help understand what results you might get. I have spent a lot of time building these models and you might, amid after a couple hours, step back and be like, "What was I doing again?" And just realize that you spent way too much time in a spreadsheet. So business growth happens outside of the spreadsheet. Just focus on executing and uses to help drive where you're going.
