So What? Marketing Analytics and Insights Live
airs every Thursday at 1 pm EST.
You can watch on YouTube Live. Be sure to subscribe and follow so you never miss an episode!
In this week’s episode of So What? we focus on your annual planning. We walk through what predictive analytics is, the use cases of predictive analytics in marketing and how to apply predictive analytics to your annual plan. Catch the replay here:
In this episode you’ll learn:
- What predictive analytics is
- The use cases of predictive analytics in marketing
- Applying predictive analytics to your annual plan
Upcoming Episodes:
Have a question or topic you’d like to see us cover? Reach out here: https://www.trustinsights.ai/resources/so-what-the-marketing-analytics-and-insights-show/
AI-Generated Transcript:
Katie Robbert 0:18
Well, hey everyone, Happy Thursday. Welcome to so what the marketing analytics and insights live show. I am Katie joined by John over here. He’s on the side today.
John Wall 0:29
Hello.
Katie Robbert 0:32
Chris is in Denver at the Digital now conferences. I think he might actually be speaking right now as we’re recording this.
John Wall 0:40
Yeah, hopefully breathing well, there’s not a lot of oxygen out there, man. He’s got it. He’s trained.
Katie Robbert 0:44
He’s, he’s trained hard for this. This week, we are talking about predictive analytics for your annual planning. We’re going to cover the basics of what Predictive analytics is the use cases of predictive analytics and marketing and applying predictive analytics to your annual plan. So, John, you’ve been with us for a long time. And it’s, it’s a type of analysis that I feel like is sort of the unsung hero of any kind of planning. What’s your take on the usefulness of predictive analytics?
John Wall 1:21
Yeah, you know, there is the challenge that it is kind of like 300 level marketing, right? Like, if you’re some, if you’re a team that doesn’t really know what the heck they’re doing, and faking their way through it for the first time. You know, it takes a long time before you can get to predictive. But once you have things in order, and you’re looking for the next step, predictive is crucially important to figure out how to effectively allocate your limited resources, because that’s always the challenge is, you know, there’s never enough, really, there’s never enough time, I mean, you can use money to buy more time, but if there’s only so much money to so you have to get as much done as quickly as possible with what you’ve got. And having predictive to point in the right direction is really the maker break. And, you know, for smaller companies, that’s whether you flame out or not in for bigger companies, the stakes are just huge. I mean, if you’re running predictive on your ad cycle, we’ve seen clients saving, you know, 10s of millions of dollars on their ad spend, because they’re not, you know, burning it that week, everybody’s at the beach, or, you know, when that big trade show is so, yeah, there’s a lot of ways you can use it, you know, there is a lot of hype about it, you know, being a crystal ball, but the reality is, there are repeatable cycles that you can find and take advantage of, and you’ve, you’ve got to be able to do it.
Katie Robbert 2:41
One of the first actually, the very first talk I ever gave on stage was at inbound in 2018. And it was like five use cases of predictive analytics, which, at the time, to me was felt very basic, but I tend to forget that I’ve been working with Chris long enough that these concepts are not new to us. But you’re absolutely right to a lot of marketers, you know, it’s, it still feels like a crystal ball, but really, it’s just using their data, doing some math and projecting it forward.
John Wall 3:15
Yeah, it’s, uh, you know, getting in and getting your own data and that you get that’s another part of it. It’s just an extension of leveraging your own data, you know, tons of organizations have this stuff already to go. And they’re just not putting it to use and squeezing the goal, though.
Katie Robbert 3:31
Well, I mean, so that’s a really good segue. So we have, we have what we’ve called the data driven marketers essential planning guide. And so what this does is this kind of walks through. So John and I are just going to use this as the foundation for our conversation about predictive analytics. And so you can grab this deck if you want it at trust insights.ai/insights/white papers slash data driven marketers planning guide, super rolls off the tongue, but will drop a link to this in the show notes as well. And so what we have done here is we have a short deck that sort of walks through what Predictive analytics is and how you can use it. For marketing campaigns, I tend to think that this is one of the better tools for email marketing campaigns, but you can use it for a variety of other things. So we will get into that. So the purpose in this case, the use case, is to tell you the good or bad times to launch marketing campaigns for the calendar year 2024. Moving into next year, and the last quarter of 2023, which is what we’re in currently. When’s a bad time when people are out of the office? People aren’t around, they’re not going to see it. To your point, John, understanding the timing and the seasonality of when people are sitting on the beach? Probably not The best time for you to go full throttle and launch your biggest campaign in the year. And then nobody’s gonna see it. Our methodology for this particular report is when they search for how to turn on their oath of office auto reply. So we’re using SEO data, we’re using Google Trends data. And those two datasets together are fairly rich. Google Trends, if you don’t know is a free tool for marketers, I’ll actually pull it up. And I believe it goes all the way back to 2024 2024. That would be forward 2004. Google Ads summary not helpful. Here we go. Google Trends. So Google Trends free to everybody, you should be using it. It’s a great tool. So let’s say we look for marketing, just as an example. pretty broad. But what you have is you can break it down by demographic region, we have all the different countries you have worldwide, you can do just the United States, you can go yep, all the way back to 2004, which is a fairly rich data set. We’ve run into terms like NFT, and other terms that just didn’t exist prior to a certain timeframe. So that’s something that as you’re searching to be aware of us, did this term or this product, or this catchphrase, or whatever it is, you’re looking for, even exist. So we can go back, you know, the past five years, always good to show some seasonality. If you work in a certain vertical in a certain industry, then you can start to break it down to I only want to know marketing in the travel industry, or I only want to know in science. For us. We are agnostic of any of those industries, we like to know all of them. And then this particular filter, I think is highly under utilized because right now we’re looking at websites, we’re looking at what people search, like when they search remarketing on a Google search. But you also have image search, you have new search, you have Google Shopping, and then YouTube search. So if you’re trying to create video content, you can see that you can see the search trends for when people are looking on YouTube specifically, which is part of the Google universe. Do you use Google Trends at all? John?
John Wall 7:29
Yeah, it’s a good eyeball for just like you said, you know, every once in awhile, you run into something like NF T. And you’re like, Okay, when did that start? And, you know, where’s the hype at? Yeah, you can’t beat it. It’s really Google that has gone out of its way to make sure that we don’t get to see much of anything. This is one of the rare exceptions where you can get a little peek, at least in relative terms of what’s going on. It’s hard. You don’t have any hard data to calculate against. But you can at least get a feel for how, like this list here is a perfect list, you know, a list of five things and how do they compare what’s on the way up? What’s on the way down? She can see AI all over the place, you know, right now dominating, unsurprisingly.
Katie Robbert 8:07
Yeah, that I was tempted to put in? Well, let’s just see what happens when I put in AI as a search term. I would imagine you’re going to see a huge ship. There it is. So basically, last summer, people started to get wind of it. Last November, October, November ChatGPT, was like, Hey, knock, knock, I’m here. And then all of a sudden whoosh, which significantly dwarfs the actual marketing and the actual advertising terms. And so again, those are just really interesting things to see this data you can download into a CSV and work with it. Super handy.
John Wall 8:46
Yeah, ChatGPT, I’d be interested because we’ve seen there was some stuff about actually user numbers dipping this month for the first time. Yeah, you can see that there. Yeah, but it’s definitely still like, it’s still weird that there was a dip there. Oh, I guess that’s just like the summertime.
Katie Robbert 9:03
Yeah, well, and what we don’t have in here, what we haven’t done is like full research on so ChatGPT, obviously, you know, picked up, picked up, picked up and then dropped. What else was happening here, what else was launching at this time that ChatGPT got overshadowed for a little bit. But that’s really helpful to start to understand once we have more than a year’s worth of data. With having a system like ChatGPT, we can start to see the seasonality. And that’s really where Predictive analytics is helpful is understanding the seasonality. And it’s not going to be the same for everybody. So we’re looking at general search terms. So this assumes, you know, B2B, B2C, all kinds of consumers, but that’s where you go back into the all categories and if you work specifically in business and industrial, for example, you actually start to see very different results marketing all of a sudden picked up, whereas that significant dip with AI and ChatGPT He no longer really exists. So they’re all the way down here now, because I switched the industry.
John Wall 10:08
Yeah, that’s interesting to see it flattened out a little bit over there.
Katie Robbert 10:12
Yeah. So if you have hobbies and leisure, for example, you’re, again, you’re going to have very different results. You know, you still have that dip, but not as significant as when you look at everything all blended together. So just a little bit of context, this is a dataset that we use a lot when doing our predictive forecasting, because it really gives you an understanding of user intent, and their search history. And so it’s a really good solid data set to anchor your predictive forecast to. So when we go back to the actual forecast that we did, you can see and here’s the example of Google Trends. So we looked at out of office, Outlook out of office, Gmail out of office past five years worldwide, because we wanted to see when people were going on vacation. And again, it’s not going to be the same for every industry, if you want to drill down even more specifically, if you want to drill down to specific regions specific states. That’s really a good idea as you’re getting into this predicted forecast for yourself. This is how predictive forecast is calculated. So John is our chief statistician. Do you want to go ahead and just like, sound it out? And tell us exactly what we’re doing here?
John Wall 11:26
Yeah, give the rundown of this model here. Now, this is yeah, this is definitely ugly math that I try to stay away from this will ruin your day.
Katie Robbert 11:36
A predictive forecast, if I recall correctly, and Chris isn’t here to correct me. So I can say whatever I want, is a traditional predictive forecast uses and a Rhema model and ESRI model for seasonality. I do remember when I presented predictive forecasting into 2018. On at inbound, I had down all the different steps, what arena stood for, and it’s, I couldn’t tell you what it is, I probably should have looked it up. But I know it’s like moving averages and seasonality. And so basically, the way that we describe it is sort of like a GPS. And so you need to know where you’re starting and where you’re going. But then the ARIMA model starts to fill in all the different pieces that are actually happening. So you have you know, whether you have traffic you have, you know, construction, other cars, and you know, timing, so you may start with, it’s a 10 minute drive. But as you start going and more data is coming into the model, it’s like, it’s no longer a 10 minute drive, now it’s a three hour drive, you’re probably better off walking.
So when so in this example, we pulled the three different datasets from Google Trends. And you can see, this is a very hard way to review it. But this basically starts to tell you as we apply the predictive forecasting model to the data, so all the historical data, what it does is it projects forward 365 days, and we can start to see some seasonality. And so you have the peaks here are when people are searching the most for going out of the office. So these would be the times when you wouldn’t want to be planning big campaigns. And these down here are the lowest meaning people are in the office. And you would want to be planning a campaign. So you can start to drill in to the individual data points to say, All right, I have a campaign planned for January 1 2025. Well, you can see in January, a lot of people are planning to not be around based on historical trends. So maybe you should either do it before or after that date.
John Wall 13:47
Yeah, and then you’ve got the peaks to jump on. Right? That’s always the thing. If you’re looking for spend to get out in front of folks, when the search volume is at its peak, you’re gonna, Well, Kristen, for this, it’s in verse four out of office, but still there’s, you know, again, you want to either go high and go low, and then the middle of you, you know, tend to ignore depending on what’s going on.
Katie Robbert 14:09
Well, and it’s interesting that you bring that up, because it really depends on what your strategy is. And so you could have a strategy where it’s actually better to launch when people aren’t around, if you’re doing like a soft launch, for example. So you want to sort of get it out there, test the waters, but you don’t want it to make a big splash. Maybe it is better for you to launch something when people are less around to sort of like let it build up some steam. The other thing to consider is what are your competitors doing at that time? So are they laying low during those times when they think nobody’s around? So it’s actually your time to strike because there’s no competition. There’s no you know, competitive search traffic. There’s no other products selling out at the same time, no campaigns to take people’s attention away. So you could actually use it to your advantage.
John Wall 15:03
Yeah, that makes sense. So you know, you keeping an eye on the rest of the spend is always that’s another poker game. But there’s yeah, there’s a lot going on on that front.
Katie Robbert 15:13
And so with this, we, you know, we do the whole analysis, what we put together is a very simplistic way we think it’s simplistic way of looking at it. And so these are all the weeks and q4. So where are we now it is November 2. So we are in this week. So we’re sort of in a, if you want to be sending email campaigns or doing any sort of paid ad campaigns, you’re in a good spot to do that. We start to get a little bit more tricky as we get towards Thanksgiving and Christmas, understandably, so people take off a lot of fun in November, December, or the holidays in the United States. And so if you’re planning on doing something, it’s probably not a great time to do it. Or if you’re like us, you sort of just do it anyway.
John Wall 16:08
It’s interesting that there’s not a huge dip around Thanksgiving. But I guess that does make sense because first of all, the rest of the world is not on board with that. And then it is a partial week, like people are definitely working. You know, most people are working Monday, Tuesday, Wednesday.
Katie Robbert 16:22
Yeah, I was looking at our calendar, and we only have Thanksgiving, which is the Thursday and the Friday marked as out of office, but the rest of the week, and then the following Monday, we’ll be around I think people tend to save that time more for the actual Christmas, Hanukkah, other holidays that happened in December.
John Wall 16:42
Yeah, if you don’t have it done by that Friday, before all the Christmas stuff starts, that’s usually it for the rest of the year, you’re thrown in the towel.
Katie Robbert 16:51
And then we can just go through the other quarters pretty quickly, because you’ll start to see some seasonality. And so you see back the second week of January, that’s a great time for that quarter. The last week of March is the worst time that you start to see in late January, and then in March. And so John, I don’t have kids, but I’m assuming that this starts to coincide with like, February vacation and spring breaks and all that kind of stuff.
John Wall 17:23
Yeah, there’s always these competing weeks, you know, there’s some schools that will do a week in February and a week in March. And then there’s usually a different set that’s often tied to a more collegiate group have taken two weeks in April. But yeah, there’s all kinds of variations with that. I know that. Yeah. See, like, it’s weird that near the end of February. That’s actually I know that that’s like a dead week at Disney World randomly, like because all the vacations fall around that people are actually working heavy that week. And then as you get to that 310 week, it gets ugly. And then yeah, I think with the four, three, you’re just like square in the middle of spring break. You’ve got both college and primary secondary schools often. So yeah, there’s people trying to get out of the cold. And there’s a lot going on as far as those weeks. So the other one is trade events to like trade events finally start to pick back up in March, you know, they go dead during December. And then January, February, everybody’s doing planning internal stuff. And then we usually see end of q1 into q2, business travel stuff picks up again.
Katie Robbert 18:30
Well, let’s see what that looks like. So this is q1, this is q2. So first, first whole month, so you have April, and then the first week in May, all good to go. At least based on historical data. People should be around. So if you’re trying to launch something, and have q1, early q2, April seems to be the way to go. Again, you know, this isn’t a specific analysis. This is fairly strict, fairly general. But we do offer the ability to do more specific analyses, if that’s something that people are interested in. unsurprising. When you start to get into MAE, you start to get into at least United States you have Memorial Day because you certainly get out of school. And then you have June in the summer months. And so June it looks like is just don’t even bother. Everybody’s like trying to wrap up school at out on vacation, get the heck out of their office, like don’t bother launching anything, then it’s a terrible time. If that’s what makes sense for your company. Yeah,
John Wall 19:33
that backup of of June, it’s graduations, and then yeah, summer vacation start to kick off and everybody’s on the road.
Katie Robbert 19:40
And so let’s see if that trend continues into straight q3. Yeah. Oh, and unsurprisingly it does. So, you have July, you have the first part of August and then mid August, people start to come back. Now depending on where you are in the United States. Some schools actually start in mid August I know my in laws live in Arizona and come like the second week of August, they’re already back in school.
John Wall 20:08
Yeah, you’ve got that seasonal thing is, it is really weird. And that’s you get people, those people bailing at the front end, too. But then it’s interesting to see that 915 week, you know, right back in action, that’s always where it’s like Hubspot inbound and Content Marketing World, and there’s like five shows that run over the next two weeks right there, like everybody is right back on the job at that point. And it Yeah, it’s interesting to see Fourth of July being the high watermark firm people bailing.
Katie Robbert 20:35
And you know, what’s interesting about a predictive forecast is, you know, and you said this earlier, John, it’s not a magic ball. It’s not, you know, tea leaves and crystals. It’s really a confirmation of the anecdotal evidence that you already have about your consumers. And so we can say, you know, we think that the week of July 4 is probably the worst week to send an email. But once we went, when we run the data, we can confirm that and then that is a decision that we can stand behind, because we can say, the data shows what we already know. So it’s a really good way to just sort of confirm what you know about your audience anecdotally, anyway.
John Wall 21:13
Yeah, and, you know, we do these kinds of analyses all the time as far as, okay, so this is just a general term, you know, based on search data, but you can run the same thing on your email performance, you know, you can look at what your how your email campaigns have done. And if you’ve got multiple years of data, you can actually see, okay, when the clicks and opens happening, and you can create, basically the same report for yourself, but it will be a lot more in line with the performance that you’re getting from your email campaigns.
Katie Robbert 21:40
Yeah, absolutely. And I think that that’s a really good takeaway is, you know, an analysis like this, like the one that we created that, again, you can get from our website is a really good starting point. But your best bet is to use your data because your data is going to be specific to your audience and your seasonality. And that’s absolutely something that we can help with. And so we started with q4 2023. So we knew that, you know, basically, December was the worst time and you can see reflected here, sort of the same thing. So you have, you know, you sort of go through October, October is pretty good. We’re in this week here, moving into next year, and we actually had the same sort of neutral response, beginning of November, because I think people are, they’re thinking about taking some time off, they can’t take too much because they want to save for for the end of the year. But, you know, attention spans are low. So as you’re planning out your campaigns, November, maybe start winding things down unless you’re in the retail industry. And then you really need to be amping things up. And then you have December, which is pretty much for B2B. Not a great time. And again, that sort of goes back to what is your industry who is your audience, so for this, if you’re if you work in retail, if you work in travel, this isn’t the best analysis for you. But if you work in B2B, and you work in tech, this is a really great analysis, because it tells you when people are paying attention, when they’re sitting physically sitting at their desks, reading emails, and looking at campaigns. And so, you know, we asked the question, what do you do with this information? Well, there’s a lot of things that you can do with it. And so, John, you were sort of talking about this a little bit. You can do general financial planning, you can do keyword topic, research, content, calendars, ad budget planning, attribution analysis, the ad budget planning is an interesting one. When Chris and I worked at the agency, we had a very large, B2C brand that we worked with. And they were spending upwards of six figures a month on paid ads, which is not an insignificant amount of money. And so in order to get really good results, we actually applied this predictive forecast technique to their keyword data, to help them tune up and tune down the spend on their ads. And it was really interesting to see it was one of my first introduction to predictive forecasting about eight years ago. in its infancy in terms of this application, and just the, the different outcomes we got when we use the forecast, and when we did was just insane. Because a lot of companies and I’m sure you can speak to this, too, John, a lot of companies are like, let’s just keep throwing money at it until something happens. Which is it’s not that’s not a strategy that’s just a panic move. But if you actually use something like a predictive forecast to go oh, now I know when marketing is trending versus when it’s not trending. Let me sort of adjust my budgets accordingly.
John Wall 24:50
Yeah, the panic Double Down is never a smart play. You know, we see that a lot where it just throwing more money on the fire and there’s just certain weeks where it’s not going to be Can a difference?
Katie Robbert 25:02
Yeah, um, you know, we talked about it in terms of content calendars as well, this is something that we’ve been using with a lot of our clients. And so if you have an editorial calendar, if you’re writing content for your website or articles, to get placements elsewhere, understanding what kind of content consumers are looking for, and when is a very powerful tool. And so with generative AI, becoming even more sophisticated, we’ve been able to get really in depth with our predictive forecast for content specifically. And so we’ll have keywords and suggested topic titles, and then all of the other pieces, and so we can go so far is to write a whole predictive calendar that has keyword, what week it should go. And then here’s a draft or an outline of the content, which is such a powerful tool, because then you just hand that over to your content writers and say, This is exactly what we need to do. And then we needed to have it posted by the state, and then that’s when people report you should theoretically see your search volume start to really go up.
John Wall 26:13
Yeah, that’s huge, you know, because pretty much anybody that’s doing any kind of content marketing has a list of the topics and things that they want to rank for. And to be able to run that through the same models and see which weeks, you know, they get the most action that can really make a huge impact and the amount of spread and organic diffusion that you get of your stuff. Because, yeah, you could have, you know, five or six topics, but if two of them are highly seasonal, there’s times of the year where it makes no sense to be writing, you know, you don’t need your pumpkin spice articles dropping in March, like that’s not going to do you any good. You know, in whatever your industry is, there’s sure enough, there’s some kind of pumpkin spice terms.
Katie Robbert 26:57
Now, I’m like, my brain is like, what are the pumpkin spice terms like for, you know, it’s basically the jargon like the digital transformation, I would call a pumpkin spice stir.
John Wall 27:08
Yeah, and for us, you know, it’s totally about, you know, when people are looking at events, or if they’ve got a product releases, you know, things like that. There’s all kinds of cyclical stuff that you’ve got to work around in lineup. And yeah, knowing what the heck is going on in the industry. And what that hidden wave is, is where you get instant lift without having to do any different work. You’re just putting in the right sequence.
Katie Robbert 27:35
Yeah, that’s a really good point. So Brian, our buddy Bryan Piper, he has a question, are you using ChatGPT advanced data analysis tool, or other AI tools to help with any predictive analytics and forecasting? I know, I can at least say, for myself, I have been playing around a lot with ChatGPT advanced data analysis tool. Historically, Chris has written his own code in R. And so it’s always been a proprietary tool for Trust Insights. That said, it’s becoming more and more accessible to marketers as ChatGPT, and other generative AI tools are evolving. Up, but I do believe that we still run ours using our own code, but the topic titles, suggested summaries those things we are using ChatGPT. For those, once we actually have the analysis done, we’re using ChatGPT, to help us take it that step farther to create the outlines and to create the things that we would hand over to the writers.
John Wall 28:47
Yeah, we have a lot of that stuff in house. But I haven’t had a chance to play with the data analysis tool. I know that that was recently rolled out, that’s something else that need to spend some time getting into where we have seen I’ve seen Andy Crestodina doing some stuff with that. And Chris has done some stuff with that, too, where you actually upload tables of your own stuff, and have it give you you know, reporting back, I don’t know as far as you know, does the thing have enough in the corpus to be able to call out seasonal trends for industries where you’re not seeing a data? That’s, that’s a different story.
Katie Robbert 29:18
I actually used ChatGPT The data analysis in this past week’s newsletter if you want to subscribe TrustInsights.ai AI slash newsletter, I use the advanced data analysis tool to help me analyze qualitative survey data feedback data to help understand what people had said on our one question survey though it actually the analysis came out pretty good. Now the to the question that Brian had about are we using ChatGPT. So you’re right. Christie had built his own tool and he is using those math formulas that we showed earlier. That is absolutely 100% What’s happening. One of the reasons for that is You know, up until recently ChatGPT didn’t have this advanced data analysis tool. So we have been building it all in house, as John said. But the other thing to keep in mind is that the, the timeline of ChatGPT, the data only goes back so far. So number one, you have to be providing the data yourself. But, you know, I, I, I’m going to try not to bungle this too hard. But basically, my understanding is that, yes, ChatGPT can do the data analysis, but it may not fully understand seasonality as its own historical database may not go back far enough, it’s maybe just looking at your data. If your data doesn’t go back far enough, if your data only goes back a year, you’re not really going to understand through seasonality. So I would say it’s worth experimenting with, to see what kind of results you can get, but know that there may be limitations for now, on what ChatGPT can do with that kind of historical data and forecasting.
John Wall 30:59
Yeah, and it’d be interesting to you know, given that it’s kind of constantly adjusting, I mean, because you could ask the same question a number of times and see how much variation you get in the answers you get, I’d be interested to see how that works.
Katie Robbert 31:12
I would say, if you’re going to ask ChatGPT, for predictive forecast without providing IT data, then you’re definitely not going to get something that you can take action on, you should at least have your own core set of data that you’re good to, to ask for assistance with that. The other thing I learned through Andy Crestodina, and I applied this, again, in the newsletter, this past week was you can ask ChatGPT to show its work. And so if you give it a dataset, and you say, this is this is what you’re doing, this is the type of analysis show your work, it actually gives you the snippets of code as goes through each step. So you could take those code snippets, put them in your own code language, like Python, I believe is what it typically writes in. And you can start to build your own models that way. I also my understanding is also that the code is usually about 90%. Correct. And so there is some error in it. So you would have to then know what you’re looking at to be able to keyway it properly, and put it into a cohesive script that you can run repeatedly. But it is a really good place to start. If you’re just building those things, you know, from scratch. And so John, that kind of brings us to our favorite part of predictive forecasting, which is the cheese of the week. And so we started doing the cheese the week in 2018, actually, and so we try to refresh this every year, because this is a really great, concrete example of what a predictive forecast could look like. And so this is Trump 2022. So it is a little bit outdated. But the basic idea is the same. And so if you were a marketer, who was trying to use a political forecast, this is how you would do it. And so you would have all of your keywords over here. And so you would want to have your keywords, not just a general set, but things that are specific to you. In this case, we’re looking at all the different kinds of cheeses. So we have monster cheddar, Swiss Parmesan, ossia, algo, parmigiano, cottage cheese, American cheese, so on and so forth. Because everybody loves cheese. And so then you have all of the different weeks. And so what we would do is we would say, alright, so for the week of October 16, what are the terms that are going to be trending the highest, so that we can, you know, over here in August, maybe start playing out our content calendar and say, what content should we have ready for the week of October 16? Well, when we sorted it, it looks like we should have content around monster and shatter and Swiss and Parmesan. And so what we might do is say, All right, do we have recipes? Do we have you know, buying guides? Do we have storage? Guys? Do we have other you know, history of where it comes from the different types, all the different kinds of helpful content that will give your consumers something to really, you know, learn about and search for so, you know, they could be like, where do I buy monster cheese? How do I use monster cheese? Can I grow monster cheese, answering all of those questions. And a predictive forecast gives you the time to do that, so that you’re not getting to the week of the 16th going on. And everyone’s looking for a monster and we don’t have any content around it.
John Wall 34:38
Yeah, this kind of stuff we had run for other food manufacturers and some other verticals. But and we were so cool when we had run it when we realized well, of course we obviously can’t share any of that stuff. So that was where cheese came into the thing. And yeah, it highlights all kinds of crazy stuff. When I’m talking about this with clients all the time. I always bring up alumi you know the fact that you see it You know, it basically spikes in the heart of the summer. And I knew nothing about it prior to this. And the fact is, it’s grilled cheese, right? So that’s July, August months, when people are firing up the grills, suddenly it gets the spike. So if you were a cheese, you know, if you ran a cheese manufacturer or your retailer that was looking for what to focus on for the next couple of months, this kind of report shows you what the cycles are. And yeah, having played around with it for a bunch of, you know, other examples, you know, about the Super Bowl, you get Monterey Jack springs up, because you’ve got everybody doing nachos. And around Thanksgiving and Christmas, you’ve got the cheddar weeks, because people are, you know, making apple pies and doing other stuff that they can bake with cheddar. So yeah, it gives you this look at your data, where, you know, when does the rest of the world search this stuff out? And when are they trying to find it? And it’s just fantastic. Because as somebody in the industry, you usually end up being blind to this kind of stuff after a couple of years working there as you don’t think about what people who know nothing about your product, like what are their first steps? And when are they looking and what inspires them to search? And this can surface all that. And yeah, and then this is just easily back solving to your content calendar. You know, it’s like, okay, you know, that homies hot July, August. So that means April, May, you’ve got to start getting the editorial going as far as writing these up and figuring out what supply going to be like, and what are the recipes that people normally dig up? And let’s make some videos showing how it works. And yeah, this kind of stuff can, again, it’s your going to do all the same amount of work, but the way you sequence it can drastically improve your results and get you where you want to be.
Katie Robbert 36:41
Yeah, we we actually we need to refresh it. We use it for Trust Insights for our own editorial calendar. Because, you know, every Monday morning, Chris and I kind of stare at each other and go, so what should we talk about on the podcast this week? What’s going on with AI? What’s going on with marketing, and we had fallen off the habit of doing it because we were running into a lot of repetitive topics. And what that tells us as marketers is we need to expand our keyword list, it was too narrow, because then you’re going to start to get every other week is something about digital marketing or every other week is something about how do I use AI in marketing, which is great. But you you get you definitely get to a point where there’s only so much content that you can create that is high quality. And so when you start to run into those issues with a predictive forecast, it means you need to start expanding your keyword list beyond the basics, and that’s where you really need to do that keyword research audience research asking people what’s going on? What do you need? What are you thinking about? What’s new? What aren’t you thinking about using tools like Google Trends to see are people even talking about this thing. Another really great way to use this is, you know, at least it used to be when you should be posting your content on social media. Social is a little bit of a dumpster fire at the moment, as Chris would say. But there are still social channels that are working. This can help you with your video content, when to post that on YouTube. This can help you when to post all your other videos on tick tock or Instagram or any of the other wide variety of social platforms out there. But essentially, you should be using this as what should I be doing today? What do people care about today? How do you know? How do I help them get to the answers they need today, it gives you the opportunity to be a little bit more proactive as opposed to reactive. Unfortunately, as marketers, we tend to be super reactive to everything, we kind of something out there and wait to see what happens. But with a creative forecast, you can really get ahead and be like this is this is my plan. And now I’m going to measure this plan.
John Wall 38:59
Yeah, and then the exception is, you know, with what we’ve been doing, and as you’re looking for topics to hit, right, there’s always black swans, it’s like what is the thing that popped up this week that we have to talk about and kick around? And of course, there’s zero history on those topics. So we have to adapt and move on the slide to accept those. But then for every other week, yeah, you should have your solid list of like, okay, if nothing crazy happens between now and then we know that this is the stuff that we want to serve up that week.
Katie Robbert 39:29
And so with that, John, you know, anything you think we didn’t cover on predictive analytics?
John Wall 39:36
You know, one huge segment that we didn’t go near at all is financial, right? If you’re any kind of SAS service or any company that’s reached a point of you’re over 10 million in revenue and you’ve got some kind of renewal cycle, you can run the same analysis on just what’s your normal buying cycle. When are people renewing when you know, you’ll get a feel for okay, we should be running renewal. Okay. campaigns on these months because, you know, two months past that is when we tend to get the normal renewals or get the churn, you know, you can do churn analysis to see what’s going on. So, yeah, it’s, you know, we talked, we showed off a great array of tools and how they work. And the important thing to remember is that these work across every front of data and every type of business that you’re doing. So same deal with, you know, who’s talking about you on social, you know, if you want to look about what are the trends for when you need to staff up to be able to handle customer service issues and things like that. We’ve done this stuff for call centers, there’s a whole bunch of other options. So yeah, no, I think we did a killer run through on what we’ve got. And just remember that, you know, this can be applied to just about any set of data that you’ve got lying around.
Katie Robbert 40:46
I think especially now, you know, with that sentiment, John, is we know marketing budgets are still being squeezed, we know that layoffs are still happening, we know that teams are overworked and burnt out. And so at least from where we sit, a predictive forecast, if you have budget to spend is a really great, you know, multipurpose tool that can tell you a lot about your business. And if your team is really stretched thin, why not help them focus in rather than feeling big, like they’re scattered all over their place? Or if you’re just bringing on contractors, you know, it gives you a really solid plan of like, okay, this is how we need to execute for the next three months, we know what we have to do, we’re not going to guess we have really tight budgets, every penny counts. So let’s make sure that we’re really using the most of what we have, and you have your data. Now you just need to project it forward in the right way so that you can make these plans and keep people focused and you know, not waste money.
John Wall 41:47
Now, that’s solid advice. That’s the way to roll.
Katie Robbert 41:50
All right. Well, I don’t know what we’re going to talk about next week. But I think that’s it for this week. If you want to join us in our free slack group analytics for marketers, route 35, almost 4000 People are in there. We were talking about predictive analytics earlier this week. There’s a question of the day every day today was about large enterprise size tools for your tech stack versus small ones. A lot of things going on in there. So definitely check us out if you want to continue the conversation if you have questions. And then otherwise, yeah, John. I think that’s it. We’ll see everybody next week.
John Wall 42:28
That sounds good.
Christopher Penn 42:32
Thanks for watching today. Be sure to subscribe to our show wherever you’re watching it. For more resources. And to learn more, check out the Trust Insights podcast at trust insights.ai/t I podcast and a weekly email newsletter at trust insights.ai/newsletter Got questions about what you saw in today’s episode. Join our free analytics for markers slack group at trust insights.ai/analytics for marketers, see you next time.
Transcribed by https://otter.ai
Need help with your marketing data and analytics? |
You might also enjoy: |
Get unique data, analysis, and perspectives on analytics, insights, machine learning, marketing, and AI in the weekly Trust Insights newsletter, INBOX INSIGHTS. Subscribe now for free; new issues every Wednesday! |
Want to learn more about data, analytics, and insights? Subscribe to In-Ear Insights, the Trust Insights podcast, with new 10-minute or less episodes every week. |