So What? Marketing Analytics and Insights Live
airs every Thursday at 1 pm EST.
In this week’s episode of So What? we focus on consumer trends. We walk through what a trend is, what data is stable and unstable, and what to do with the trend. Catch the replay here:
In this episode you’ll learn:
- When you don’t need predictive analytics
- The difference between stable and unstable trends
- How to navigate consumer trends
- Solutions in search of problems – 8/19/2021
- Email stats
- LinkedIn algorithms
Have a question or topic you’d like to see us cover? Reach out here: https://www.trustinsights.ai/insights/so-what-the-marketing-analytics-and-insights-show/
Katie Robbert 0:26
All right, here we are, again, let’s see if I can not butcher the name of the show this week’s Welcome to so what. So what the marketing analytics and insights live show it helps when I put the banner up and I can just read it this week, Chris John and I are going to cover consumer trends stable and unstable predictive data. So it’s the topic that Chris loves, because I hate it. And I hate it because Chris loves it, because he was talking about pumpkin spice. And so I was like, I can’t, it’s only August. And that’s when the crux of the conversation came about was that the way in which consumer, the consumer industry works is that you’re generally like two, three months ahead of when something is bound to hit, you start the marketing, you start the promotions. And so that’s what we want to talk about today. So in today’s episode, we’re going to cover when you don’t need predictive analytics, the difference between stable and unstable trends. And that’s a specific that’s particularly relevant as especially to what’s been going on in the past 18 months, and how to navigate consumer trends. So Chris, where do you want to start?
Christopher Penn 1:40
Let’s start with this. And I actually open up to both of you, how do you how would you define a trend?
Katie Robbert 1:55
I think if I take the opposite definition, it’ll be easier. So it’s not an anomaly, which is typically like a one off spike in something. But a trend is something if you think about it in terms of your data, or even in terms of like people think of like, Oh, these genes are trending. It’s something that more than one person is doing for a specified amount of time. And then it either drops off, or it picks up even more and becomes just sort of the status quo.
Christopher Penn 2:29
John, what about you?
John Wall 2:30
Yeah, that’s it’s interesting, because there’s kind of there’s definitely two ways you can go with it, right? One is just the mathematical definition of it’s something that has momentum and is moving in a specific direction, until it hits an inflection point. But then there’s the whole Yeah, kind of what the society do things that are that catch the Zaid guys really horrible 10 cent word, but it really is the one that fits, it’s just kind of, you know, people find out about something and it gains momentum in the press, it’s usually being searched for more on Google, or people are buying in stores or whatever. But it’s also getting that kind of upward thing. And it’s funny that when we say trend, we assume it’s always upward unless you specifically say a downward trend. We’re always assuming it’s in the positive direction.
Christopher Penn 3:13
That That is true. That is true. Oh, no, I think those are both great direct definitions, because a trend really is a general direction, which something’s changing, right? The the number of people, you know, with wallets has been on a downward trend since the 1980s.
John Wall 3:37
Like we said,
Christopher Penn 3:40
and trends are, again, those things where something is developing in some direction. And there’s different kinds of trends. We talk a lot about trends when we’re talking about marketing, because we’re trying to predict a trend and And ideally, get ahead of it so that you’re we’re prepared, we’re not scrambling the last minute. And where a lot of marketers run into trouble is, frankly, just not being able to do that. Now, there are some cases where you don’t need predictive trend analysis, which is the ability to take existing data and forecast report, anything that is rescheduled, there’s not a whole lot of thinking that needs to go into that, right. So you don’t need any kind of analytics to know when Christmas is going to be this year. Right? It’s always the same December 25. You don’t need a whole lot of predictive analytics to know when new users and Mother’s Day total to 2022 or Mother’s Day 2075 you will know because the calendar has those set dates that you don’t need forecasting but there’s other things that happen in the lead up to those things, or that are, you know, seasonal and cyclical that don’t have quite as firm dates but are still predictable. Now, where you do need analytics is when you want to forecast out, though Other things that are not literally on the existing calendar. But I think it’s, it’s challenging for a lot of marketers to, to know different types of data. So a good trend, meaning one that you can work with is seasonal and cyclical. Right. You know, generally speaking, you know, if you’re in B2B sales that you’re going to have a harder time in the Northern Hemisphere between Memorial Day and Labor Day. It’s, there’s people on vacation, you know, the opposite is true. In the southern hemisphere, you know, the December yo from November through through January. really tough time because it’s it’s summertime down there. That stuff is there’s this these trends that are all over, that we can pick out, but it has to have some seasonality, a trend that does not have seasonality cyclicality. It’s really hard to predict, like, say, global pandemics, that’s a bit. Although there’s a bit of a, like, 100 year trend on those. It’s something that we can forecast really well.
Katie Robbert 6:00
Well, and I you know, and you’re using that extreme example, but you know, there are other examples. Um, you know, think about so obviously, you mentioned Christmas time. And so we know when Christmas in the United States happens, it’s December 25, every year. But layered on top of that are the consumer trends of what’s the hot toy this year. You know, what’s the, you know, if you think about wedding season, wedding season is roughly made a September in the United States because of what it’s like outdoors. But what are the trends of the types of weddings like rustic weddings, outdoor weddings, elegant what is what are the colors? And so those are the things that you do need that data to help predict. Because, yes, you know, the seasonality and the cyclicality, but you don’t know the specifics around, you know, what people are actually going to want during those times. And so, you know, we started off with, you know, the whole pumpkin spice latte, and so are the pumpkin spice everything. And so that seasonality and cyclicality usually happens in fall, whoa, with weather changing, and the way that the literal seasons, the weather seasons are happening, you know, that timing is, you know, shifting a little bit. And so that might be where you want to use some of that predictive data to figure out when are people asking about that stuff versus just assuming? Well, it’s for the first day of fall is, you know, September 23. So I have to have all my pumpkin spice stuff ready for that day?
Christopher Penn 7:35
Exactly. Right. So let’s take a look at some examples of stable, non stable trends. I’m gonna pull up just good old fashioned Google Trends here. We’re not going to do anything fancy just yet. And what we’ve got here are three different we’ve got Christmas gift ideas, which is, you can see the blue line, really obvious trends, right? If there’s a clear spike that begins. Let’s take a look here on our map we have you know, October 22, is when that you sort of hit that inflection point. That was 2018. You hit that inflection point right around October 14 and 2019 2020. October 13. I’m sorry, 2019. October last year, October 2018, to 24th and 2020. There’s a very clear cyclicality and seasonality to it. Same for pumpkin spice, right? It’s pumpkin spice time. We can see that as of right now. You know, this week pumpkin spice is has hit its inflection point and is rapidly on the way up.
Katie Robbert 8:30
That is so disappointing.
John Wall 8:33
I saw it v J’s this week and I was going to take a picture and send it but I knew Katie would just get angry so I didn’t do
Unknown Speaker 8:39
Katie Robbert 8:41
I get angry because I think it’s gross. It’s not something that I personally like the flavor combination of. And so the fact that it’s like seeps into every consumer good. Every like, I can buy punk pumpkin spice dog food for my dog. Like Mike doesn’t care my dogs like, oh, you’re feeding me. That’s great. I don’t care what it tastes like. And so that’s why I get so angry about it. In case people were wondering because I think that the flavor combination describes
Christopher Penn 9:08
they’re actually the groomer near us has pumpkin spice body spray for dogs. Yep, that’s gross. You could talk smack that was a joke. But there was a car dealership that advertised pumpkin spice brake pads.
Katie Robbert 9:23
Oh, like. So back to the point of this not my own, you know, ranting and you know, whatever’s what you’re showing, Chris is you haven’t done any kind of AI or machine learning or data manipulation. But what you can see is if you don’t have your Christmas gift ideas, lists put together and published by the middle of October, you’re going to miss the consumers who want that information.
Christopher Penn 9:56
Exactly. If you have not already put out your pumpkin spice stuff. Is your a week late? The trend has already begun. Now what a contrast is the line in red, which is clubhouse, the clubhouse app specifically clubhouse app. And what we see here, this is an unstable trend, right? It was kind of a flash in the pan, it happened there was a lot of volume around, it was sort of a second wind and then gone. And so when we talk about forecasting, one of the most challenging things about forecasting is when you have something that’s unstable, that is itself in its entirety, and anomaly. Now, I’m going to zoom in here just to the past 12 months. There were miniature trends along the way, right? Obviously, your Christmas last year was when clubhouse really began to pick up steam, and then it kind of just sort of fell off a second wind in February and then gone. If you’ve been trying to do predictive analytics based on only data you had at this point right here, you might have predicted that this thing was going to go up forever, right? Because if you pretend that there was no other data at that point, your software would have said, Okay, well, ice clues is an inflection point here. And there’s no it hits hits over logistic boundary right there around that 50% Search interest on sale. Yeah, that’s about like a maximum that you’re going to get. And so when it comes to unstable trends, it’s very hard to predict them with any level of accuracy. We ran into some trouble with one of our clients yesterday, actually, last night, and you know, they were looking at some data from some of their customers, and saying, Wow, some of these forecasts are really off. And we’re like, yeah, they’re really off. Because you have unstable data, you have data that is not seasonal, is not cyclical. There’s no seasonality that can be detected. And there’s some so many anomalies and weirdnesses in them that it’s really not suitable for forecasting, at least not in, not with the conventional tools that are available to us. So one of the first things to do anytime you’re starting to think about trend analysis is go to a tool, you know, like Google Trends, for example. And, you know, pop it into a couple year mode and see, is there anything that indicates some seasonality? It doesn’t have to be rigid like this, but it should be something there. Let’s look at what’s let’s clear some space here. Katie, why don’t you pick something that is common, but you know, might not be as seasonal as like, pumpkin spice?
Katie Robbert 12:27
Um, we were, you know, it’s funny, we were talking about this a few months back, I would be interested to see if there’s any seasonality for skinny jeans, or any trends, really.
Christopher Penn 12:43
Five Year view. It’s not as clear to see, but there is some level of seasonality to it, right. It’s not perfect. It’s not Christmas gift ideas. But there is something there, there’s enough there that you could start to say, Okay, it looks like you know, there’s ABS inflows and this makes logical sense. There’s only certain times the year that you probably want to wait, any kind of jeans, right. You know, if you’re in Arizona, it’s the winter months is when you would wear jeans, if you’re in New England, it’s you know, the the spring, the fall in the winter, all over the winter, you might wear like, three layers of Jean. Got the kind of winter we’re having. So.
Katie Robbert 13:23
And if your name is Katie, you wear them. 365 24 seven. I mean, you do you. But I think what’s interesting is, you know, one of the things that you know, people talk about when they think of consumer trends is, is that cyclicality of well, it was popular in the 80s. Everything that was in the 80s is coming back, you know, 10 1520 years later, the 90s are back the 2000s or back, but like what does that mean? And do we have the data to support that?
Christopher Penn 13:56
That’s a really good question. Some things types of data you do have. So like Google Trends itself goes back to 2004. Right, and you can clearly see, going really far back there. Those ebbs and flows are a lot more visible right now. Google Books goes back to the 1600s. I think, at monthly level resolution, you can just look for things that show up in published books over the centuries, that’s a really fun way to By the way, look for language, it’s kind of making a comeback to terms that have fallen out of favor for a century or two. And there are these ways that we have some level of accuracy data at multi generational data points. But even you know, Google itself really didn’t start publishing this data until 2004. And Google has been in existence since what 98 Hmm.
Katie Robbert 14:47
So do you want to know the friendzone?
John Wall 14:50
Yeah, well, I get to see if we need to get in front of it. Let’s throw mullets up there on the big board.
Yes, I gotta start growing my hair baby. It’s starting to make
Katie Robbert 15:09
oh my goodness well heard on your camera anymore.
John Wall 15:18
In the front party in the back that’s
Christopher Penn 15:24
another example of actually one that’s that’s relevant to all of us is podcasting. podcasting has been around for you know, it started in the in the early 2000s. You can see here that in terms of searches for it, it has actually, you know, that was sort of the peak time, whatever is right here, what the heck this thing was, and then it has kind of nicely leveled out since then.
Katie Robbert 15:44
You know, we, we anecdotally joked, and we obviously did the same thing. But during the pandemic, like last summer, it felt like everybody had a podcast, everybody started a podcast, a live stream, YouTube show, whatever the thing was. And so it’s interesting to not see that more clearly reflected in this trend.
Christopher Penn 16:10
Again, the challenge with some of this data in particular, because this is Google Trends data is that it’s based on what people search for. So podcasting itself is a very, very broad term, you’d be better off looking for a specific show or personality or something to see if that is more reflective, like Joe Rogan, for example, would be a name that you could type in and see, because podcasting is kind of an unusual creature, where the genre overall is not as well known as individual players within the genre. Like we know, for example, Netflix is a genre of their individual shows that do really well on it, but the whole tends to be the sum of the parts podcasting is actually weird in that way, and that individual person has outshine the medium itself.
Katie Robbert 16:54
So as a marketer, what do I What do I do? Do I start with, let’s say, I’m tasked with promoting pumpkin spice, pumpkin spice everything. But I don’t know, when people want this information. So what do I do?
Christopher Penn 17:12
So there’s three steps you need to take. The first is, you’re going to need the data and probably some kind of forecast to let you know when the thing you’re interested in is likely to happen. So with pumpkin spice, let’s go ahead and see which one of these is the largest term here. Looks like? Yeah, pumpkin spice, the parent term is the, the term to pay attention to. So let’s take a look at pumpkin spice itself. The vertical line here is today, that’s where we are right now. So the point that for as a marketer, you want to be in front of a trend is ideally you want to have your stuff in market, right at that inflection point where the slope changes rapidly, so that you your stuff has time to get attention, right. So ideally, if this if pumpkin spice is something that you care about, you probably should have had your blog posts and anything needs to be indexed in search, you know, up about a month ago would have been would good. If you were doing consumer publications like magazines, depending on whether it’s a short lead or long lead magazine shortly magazine, it’s about the same as search a month out long lead magazines, like lifestyle publications can be up to four months out. So you would have wanted to be doing your pitching for pumpkin spice things in April of this year. Right now, it is much too late to do anything except swipe a credit card and run a whole bunch of ads, right? Because you were you’re behind the eight ball and on your timing. Now, that’s not to say that pumpkin spice won’t be of interest to consumers, right? It will be of interest to consumers through the end of October. So there’s a lot of volume to be had. But if you wanted to be first to get that mindshare, you know, that window has kind of closed.
Katie Robbert 19:01
Right, sorry. So basically, if I’m a marketer, and I’m looking at this and saying, Oh, we haven’t hit that, you know, peak point yet. My instincts like okay, I haven’t missed it yet. But what you’re saying is, don’t wait for that spike, to get your stuff out there, get it out there so that it’s available for people to consume by the time that the data spikes.
Christopher Penn 19:28
Exactly. Even with social media stuff, you know, people say like publish, you know, publish when when the trend is hot. You cannot trust social media algorithms to for visibility on content, even you know when it’s when you’re creating a trend on it in a timely manner. So earlier this week, we’re doing some promotional stuff for one of our clients publishing some stuff on LinkedIn and Twitter and it was designed to go on the day of launch and did everything went out exactly on time. But it’s picking up volume today, two days after launch. Because of the nature of the algorithm, surfacing things as more and more people interact with it, so even something like pumpkin spice, you might want to have seeded content out there like a week early, and then try to start creating some of that momentum on your social posts, you know, two or three days before your target date, because you don’t know what the algorithm is going to do with it or not. Hmm. Now, if we’re looking ahead, let’s take gift ideas, right gift ideas, pretty easy one for the holidays. We can see here last year, gift ideas, your inflection point really is you in that first couple of weeks, October this year. It’s good. It looks good a slightly later, so second, and third week of October, right? So hey, if you’re a consumer marketer, looking at this stuff, you’ve got a slight bit more time. But here’s the challenge. The second part of how to navigate a consumer trend is that you need to have the context, we need market research. And there’s still enough time now to do some market research to ask people, you know, hey, the holidays are coming up, what kinds of things are you thinking about? What kinds of gift ideas are of interest to you what’s caught your eye, you’re not going to be able to do stuff that’s purely anomaly based, like whatever the hot toy is going to be this year. But there’s some general leanings you can get towards, you know, last year it was people buying gift certificates or things for vacations, because like, hopefully, the pandemic called the Copa Zune, and they can go on vacation, but that market research, focus groups, consumer surveys and things, it’s really important to get the context of a trend.
Katie Robbert 21:42
What’s interesting is, you know, so the general term gift ideas, is what is giving you that trend from October to November. But you haven’t clarified this Christmas gift ideas. And so, you know, there’s other holidays or other occasions that you would want gift ideas.
Christopher Penn 22:08
You can see right here in May of this year, there is a spike in that, of course, it’s Mother’s Day, right? Smaller spike, cuz for Father’s Day, because that’s just the way things are. You see some stuff happening here just before Valentine’s Day. So there are other points through the year with where that occurs. And you can see more of that. yells, move that filter, let’s look at your gift ideas for men. And gift ideas for women are those two lines, they’re they’re pretty much in sync except for certain points, right? So there’s that there’s a, they fall out of sync during the summertime, and then kind of come back together. What I find very interesting is that the gift ideas for women peaks earlier in the fall. And and consistency is higher in volume than the gift ideas for men, right? There’s a there’s a change in volume between those two things that remains separate. It could be because speaking as as a heterosexual male, I have no idea what to get I sometimes Google for that a lot more. Like, what’s exactly, that that’s maybe a behavior thing. But that’s where that context from market research is so important.
Katie Robbert 23:20
It what occurs to me is, so we are looking at data from Google search. And so it you know, as you were talking through these different major holidays, you know, and gifts women, I was like, Oh, well, if I was, you know, if I owned a jewelry store, I would make assumptions around you know, when people were looking for gifts of jewelry, or engagement rings, or those kinds of things based on the seasonality. But wouldn’t it make more sense if I was using my own sales data? To predict or? Or could I blend the two together so people looking for diamond rings and my sales data?
Christopher Penn 24:08
I would definitely do the two together. And the reason for that is just because your sales data is offset or slightly different. doesn’t mean you’re good at selling that thing. So if you’re really bad at selling, you know, watches, for example, and you’re searching for watches, you might be have some skewed sales data because you’re bad at selling watches. It’s kind of the reason why we tell people you know, be very careful when you do things like you know best time to tweet or or whatever, based on your own data because if you only tweeted 8am your best time to tweet basically it is always going to be 8am
Katie Robbert 24:45
it So another example of this so my husband works at a major grocery store. He works in the meat department and their ordering is based on historical trends. You Usually that day, the previous year, well, you can probably imagine that during the pandemic they’re ordering and their sales was all over the map all kinds of messed up. Here’s the staff we’re going to need for a day like this. Here’s the product we’re going to need for a day like this. Here are the sales quotas that we think we’re going to hit on a day like this. And it was all kinds of wonky and messed up because it was using the summer before which for all intents and purposes was a completely normal summer.
Christopher Penn 25:32
Exactly right. And so things like the pandemic, add instability into every single trend, there’s not a single trend has not been affected in some way. Because it’s a global events like a World War, right? There’s absolutely no way. Unless you’re like that one island off the coast, I forget what the coast of India where they just like shoot at you, if you even approach unless you live there, and there’s no pandemic, they’re everywhere else has been affected in some fashion. So you’ve got to take that into account. So if you find that your forecasts are unstable, your next best bet is do exactly what we talked about Katie, which is flip over to real time data, and try to do short before casting a seven day window 14 day window, as opposed to using a year’s worth of data because the world that you’re forecasting from might not exist even today. You know, today is different, no different than a month ago in terms of like what you’re allowed to do and what you’re whether you should be wearing a mask or not. Things. This pandemic is throwing lots of curveballs. One more example, one of our favorites is to look for, when people are typing outlook out of office, we need to put that onto a different grid here. Paint it axis, it’s fixed. And to evoke that
Katie Robbert 27:03
we have real time, Chris breaking things perfectly. There we go.
Christopher Penn 27:14
So outlook out of office, when people are typing this search phrase, it’s generally because they’re trying to find the out of office mechanisms so they can go on and go on vacation and things and get out of where they are. So we see here, as you start getting to the summertime period, particularly around, you know, just before Memorial Day, throughout the summer, people are searching for this a lot, we’re actually going into the sort of the downturn of it now. And you see really hit a low there on September 12. That’s, of course after the Labor Day holiday here in the United States. But people are back in the office. And then again, once you get to November 28, which is Thanksgiving week, very likely that people are gone for the rest of the year. So if you are in B2B, this is still a consumer behavior, but it’s on the B2B side. You’d better get your your big q4 push out, you know, I’d say like November One is if you’re not if you haven’t gotten your big q4 push by November one. It’s gonna be good for you. It’s not gonna be a Merry Christmas.
Katie Robbert 28:14
Take notes. John, you and your mullet.
John Wall 28:17
This is that heavier moment ready before December?
Katie Robbert 28:20
So I had to. I have two questions for you, Chris. One is about correlation. And the other one, which I’m now forgetting. Of course, correlation. One, let’s let’s tackle the correlation question because now I forgotten what the other one was. Um, so we make assumptions around what’s going on during this outlook out of office. So what of my assumption is that the out of office is tied to school, for example. So what it makes sense to look at out of office and back to school, in the same chart to see if when one goes up, one goes down.
Christopher Penn 29:02
You absolutely could do that. You could pull in that data, you have a search data, or if you have other forms of data about that and pull it out and see if there is a relationship between those two terms. The thing you’re really looking for this what it was called cross correlation, which is, is there a time lag? So for example, if people are going on vacation? Is it reliable that two months after this session for back to school, you know, it’s there’s there’s a distinct possibility for that, at least for that certain point a year. But other times of the year, that might not be the case, right? It might be you know, when you look at Christmas, and New Year’s people don’t really talk about back to school, something not at the same level that they would you know, in September,
Katie Robbert 29:41
so, okay, so that’s an interesting way to think about it. I guess the other one that you might want to layer depending on the industry that you’re in is you know, people searching for travel. So if out of office is spiking, you could make assumptions that two or three months ahead of those trends people have been looking for Travel. Oh, that reminds me of what my other question was. So one of the things that we get asked a lot, especially now is, so let’s say I have five or six years of historical data that I want to do forecasting on. Knowing that last year was very much anomalous. Do I just cut it out altogether? And look at the past six years minus 2020? Or do I include it? And how do I account for the drastic differences in that data?
Christopher Penn 30:33
It depends, it depends on how badly affect the data was, for example, Gift Ideas was not substantially affected. There was a change for sure, yeah. But it was not so different that it was, you know, totally broken. ruses like, you know, cruise packages substantially affected. Let’s go back to Google Trends here. Let’s look at, you know, whose deals you can see this, you know, as sort of as a seasonality then gone, right. And so now, you have a question like, which data should you use? Should you use the post pandemic data, which is reflects the reality, she used the pre pandemic data, which reflects sort of the ideal state? What do you do here? This is the case, again, where we would say, I lean more into the real time data, because even now, it’s not clear what’s going to happen. There’s enough people that are concerned about things like you know, the Delta variant of the lambda variant that it may be dampening consumer interest. And so, this is a case where this goes back to our stable versus unstable, this is highly unstable, which means I would not try to forecast this. You can, you can try. But if so I would keep it a very narrow window.
Katie Robbert 31:56
So if someone came to us and said, you know, what can I do? I think that the, the appropriate response there would be, well, you haven’t gone through the past 18 months and sort of seeing different things happen, your best bet is probably to prepare for a couple of different scenarios, and take it day by day, week by week and push out the type of marketing tactics that are appropriate to whatever scenario is currently happening.
Christopher Penn 32:27
Definitely that and again, using that market research, you know, having a panel of consumers you can talk to on a regular frequent basis. It’s like, Hey, what are your vacation plans now? Right? It’s August 12. You know, you live in, say, let’s say, Florida, you’re in a state where 90% of the ICU beds are full. What are your plans? Now you’re going to Disneyworld like? Probably not, or maybe you are, but you need to have that data. And that really, actually brings to the third and most important part of consumer trends at work, which is you have to do something with the data. trend data in particular is very susceptible to analysis paralysis, where people just sit there and just kind of go down rabbit holes, like we do that, too. That’s because it’s fun. But at the end of the day, you got to do something, you know, we can look at pumpkin spice data all day, just admire it. Sit here and
Katie Robbert 33:24
I mean, there’s two kinds of people. Exactly.
Christopher Penn 33:28
Let’s see pumpkin spice cake. lattes recipes. But you got to do something with it. Because if you don’t do something with it, then it there was no point.
Katie Robbert 33:40
And so I think that I mean, that’s true of any analysis is, why are you doing it in the first place if you don’t plan on making some decisions or taking action with it. And so for us, you know, we like to joke about the unused fullness of like a pumpkin spice, you know, analysis. But what we’re able to do is use it as an example of what to do with consumer trends like we’re doing on this live stream, or when we’re talking and counseling clients around using predictive data. Pumpkin spice, for better or for worse is actually a really great example because it does have such strong seasonality. And it’s not a brand new thing like clubhouse that doesn’t have enough historical data to see, is it going to come back?
Christopher Penn 34:28
Exactly. And this inflection point here is, I think, one of the most important points in the trend because you use that that is your decision date. Actually, that’s that’s what your drop dead date, your decision date really is here. When you first see the trend beginning. That’s the point we say, are we going to do something about this or not? Right. So if if you were We were having a marketing meeting, you know, month and a half ago, you would have said to me, we need to make a decision about what we’re doing for pumpkin spice season. You know, by July 4, is it a go no go we’re going to promoted heavily are not the inflection, this inflection point here where the curve really takes off, that’s your production date, you know, stuff has to be doing getting going. And now, you know, you’re behind the eight ball. So when we look at, you know, gift ideas, right? Our I would say your, your, your production date is October 10. And then your your go, No Go is September 12, right, you’ve got to, you’ve got to have everything locked and loaded, in terms of what you’re going to spend and how you’re going to do it by the week of September 12. So that you have a month of production time to get ready to end. And you can hit that trend as it rises, which means you have to make a decision, you can’t analyze it forever.
Katie Robbert 35:44
Out of curiosity, why September 12, and not, you know, June 20.
Christopher Penn 35:53
Because at least for consumer marketing, it’s been my experience dealing with consumer brands that about a month of lead time is all you can get people’s attention on before too soon, and they just lose track of it.
Katie Robbert 36:06
Well, I guess cuz I’m looking at your, the graph that you have. And so right, there’s, you know, a dip in a spike in a dip in a spike. So how does someone who is an as well versed in the seasonality or even as well versed in reading these charts, as you are? Know what dates to set based on what they’re looking at? Because you could look at this and say, well, there’s lots of lots of, you know, you know, ups and downs, how do I pick the go? No Go dates.
Christopher Penn 36:34
So it goes to two things. One, there’s the math itself. And what is the actual inflection point, if you were to actually smooth this out, you would see those two points a little more clearly, this series here is is one smaller slope. That’s the mathematical side. But the other side is the organizational behavior side, knowing what what your company is, and is not capable of some companies, you know, for example, like Trust Insights, where a three person company like this is literally the entire company here. So if we know that October 10, is the thing, because there’s not a huge chain of command, we could do something two weeks before and say, Okay, yeah, we can make a decision two weeks before the thing. Other companies, it might be six months, it might be almost a year, you’re saying like, yeah, we budget approval for the 2020 on a holiday season, you know, in like January one, we got to have your budgets and so you got to know your company.
Katie Robbert 37:22
Makes sense. So that goes back to your point of the data is great. But without context, it’s really hard to do anything with it, because then it’s just, you know, wines on the screen.
Christopher Penn 37:32
Exactly. What is the customer going to do? And what are you capable of as a company. And those two things have got to agree, we’ve got to be in sync, if they’re not, you risk being that company that it’s you know, December 1, and you’re you’re 40% behind your sales goals, and you’re like, Okay, sell everything. You’ll feel left hand, this book shifters 80% off, and you’re like, no, it’s, it’s just really bad. Or you could also be the company is like, we just need more data. We need more data, we need more data about the 2014 holiday season. It’s January 1 2022. Like, we finished the analysis like, right.
Katie Robbert 38:08
I it’s weird, because you’re describing things that I might have had, like nightmares about, or maybe have lived through in other lives.
Christopher Penn 38:17
Exactly. But that’s, that’s the, the heart of this. So, you know, in terms of, of what we’ve talked about today, and you know, what is a trend when you don’t need to predict predictive analytics. We’ve seen stable and unstable trends, but it really comes down to you need the trend in the forecast, if you can get it if it’s stable enough to forecast, you need the market research and your organizational reports for that context. He got to make a decision.
Katie Robbert 38:44
So the big takeaway is, if nothing else, if you have no other resources, Google Trends is free. Start there. Ask questions to Google Trends, put in a bunch of different keywords, use your SEO keyword research or your keyword list and start plugging it into Google Trends and see what comes up.
Christopher Penn 39:04
Exactly right. So yeah, that’s that is sort of what to do with trends. If you have some pumpkin spice recipes, come on over to TrustInsights.ai dot AI slash slack
Katie Robbert 39:17
group everyone’s getting there’s
Christopher Penn 39:20
a food and drink channel the
John Wall 39:22
food and drink thing if you know it’s good for you. We’ll
Christopher Penn 39:28
talk to you next week. Take care. 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 slash ti podcast and a weekly email newsletter at Trust insights.ai slash newsletter. got questions about what you saw on today’s episode. Join our free analytics for markers slack group at Trust insights.ai slash 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, Data in the Headlights. 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.