Forecasting Perils: Where Things Go Right and Wrong

Will:
All right, ladies and gentlemen, we are back with another episode of Safety Sock. I’m Will Davis. And staring at me straight ahead is Dan Magida. Dan, what’s up?

Dan:
Not too much man. How, How you doing?

Will:
I’m doing all right. How’s California?

Dan:
It’s good. I will say they’re not wrong when the sun is always out. It is. It is quite glorious on how good you feel from relaxing in the sun.

Will:
Well, I understand your need to relax and Dan, that kind of actually brings us to our topic today. And when companies are doing well, it typically means that their goods are arriving on time and that their teams are getting the appropriate forecast to know how much goods they should order in the future. But a lot of those teams don’t get accurate forecasts and sometimes they don’t even get forecasts at all. And I think, damn, we’re gonna jump into that subject and understand why is it such an issue for companies to create accurate forecasts for the goods that they purchase?

Dan:
Yeah. So are they relaxed or are they not relaxed? I guess it depends who you ask on the company or the suppliers for that sake. If they feel relaxed.

Will:
Yeah. I feel like the ones that are doing well are the, are the ones who are relaxed and the ones who aren’t are the ones that are stressing out about it.

Dan:
Yeah, there’s, if you’re doing this correctly, there’s a lot of planning that goes into a building, a forecast and multiple teams that you’re communicating with. But also looking at your, the big thing, your customer data and what your retention rates are that are driving not only new business but those recurring customers. Cause that’s what makes up most of your base. And then making sure that they have enough stock to head into next year. So we’ll, we’ll dive into that and talk about what systems work best for you guys to, to capture all that.

Will:
Yeah, so Dan, you touched on it that there’s really like three main areas where things go wrong. You know, one could be the data you’re using in order to create your forecast. Another area that can lead things. Soray can be the communication between the teams that are ultimately building the forecast. And then lastly, you know, what’s the process in place that teams have in order to generate it and are they doing it in time? And I think we can actually come up and share three good examples of when these have gone wrong and, you know, at points in time when things are going well. So, Dan, which one do you wanna start out on?

Dan:
No, we can start from the top. Let’s just take it from the, let’s take it from your top down approach there.

Will:
All right, let’s do, do it. So yeah, so you know, from a company side, Dan, I think where a lot of people struggle is not that they don’t have data, but they’re not necessarily sure what to do with the data once they have it. And what’s the best way to accumulate the data and then put it into a place where they can action on it. Have you seen this before?

Dan:
Yeah, I mean I’ve had experience forecasting and there’s, I mean, you have a lot of different outputs. I think the main thing you need to understand is where you tracking your sales. So what’s your point of sales system? Are you strictly direct to consumer? Do you have a retail outlet? Just understanding where your end consumer is buying those products. And then mapping out that customer journey. How many new customers per month did you have? Cause that feeds into your marketing forecasts. And then of those customers who bought previously recurring, how often are they buying each month? So then you know what your output is for total units sold. And then you can also tie this into your fulfillment data as well. Cause a, everybody has a fulfillment center of some sort and that should be giving you accurate reads on what’s leaving fulfillment center on a daily basis. And then just bring that into your model. You can trim this up on a monthly basis. It could, this can go into your, say a SQL database or you can put this in Excel file or we’ve talked about other systems like inventory management systems where you could store this data as well. But you see a centralized place for storing all of the outputs that you could be

Will:
Gathering. So Dan, you know, for brands that are starting out or you know, getting bigger, this is where companies like Shopify can help them right from point of sales perspective and also aggregating that data to push to another location.

Dan:
I mean the, they us check out system for sure that feuds into their fulfillment center. Now I don’t know how complicated a matter they are if they’re more one off or recurring, but then they, there’s also Shopify has outlets that you can do current subscriptions from as well, that plugin. But yeah, you can use any of those point of sale systems that feed into your fulfillment center just to get the more specific customer cohorts or you can just operate off of you just your fulfillment center. And just know I sold 500 units in January, February was 750 and grow the multiplier from there as well. Putting different ways to dice it up. Also know that bandwidth is an issue as well in those early stages.

Will:
Yeah, so then that, that kind of takes us into the second point. And this is where you see a lot of bigger companies where they struggle is how the information is communicated amongst multiple teams. And you typically for larger companies, have a certain breakdown related to marketing who is choosing the direction of what something’s going to be or something that’s going to get launched. And then you then have sales people who are going out and selling, you know, what the marketing team is saying you’re going to have. And then that sales team is coming back to a demand planning group and sharing the data saying, Hey, I’m going to buy 200 of this. I’m going to buy, you know, a thousand of this. And that demand planning team is then aggregating that information, taking a look at it historical between ups and downs. Are they going to add 10%, 20%, you know, seasonality, Will they take 10% away? And then, then they are pushing that data to a ordering group who then is responsible for ordering those goods. Now Dan, I’ve seen it at companies where this works really well and it’s a pretty seamless process and the communication’s great, but when there’s a breakdown here, it really screws things up. Have you seen something like that before?

Dan:
Yeah. So like, this is not to go too technical here, but there’s those early stage companies. What you’re really focused on is hyper growth. And, and that could be cause of a, you’re trying to prove market fit and your backers, if it’s venture capitalists or just private money, wanna see growth. So you get very aggressive with your growth and start throwing out crazy projections that are, that are gonna impact not only spend on your supply side, but also spend on the marketing side as well. And then that just burns cash as well. So the approach that I’ve taken the past, and this was more at a company that had a recurring subscription model, was, you know, who your customers that are ordering each month model that out and then you know what your net new is solve for the recurring base. Cause that’s your loyal customer base and make sure you have enough supply to reach them and then fulfill the new, The one nice thing about depending on your model, I would say in supply chain is the MOQ model.
There are pretty substantial jumps at times. And the one difference between demand forecasting, supply forecasting is supply forecasting in you’re ordering based upon if, if it’s MOQs to start, but then on truckload sides in the US or container sizes just for optimizing freight. So you can take those numbers and then convert it into more healthier metric on the supply side to support a those bigger buys. I know will, you’ve, you’ve done a bunch of that as work as well where you take the demand, whatever those projections are, crazy or not crazy, but then convert it into more of the supply side, which is more reasonable to look at because your max you’re optimizing across multiple different price sectors.

Will:
Yeah, I I think when you take that into account, but also comes to mind is, you know who you’re partnering with on the supplier side because as you grow, you’re going to continue to ask your supplier to be flexible with you as you, you know, go into new journeys from a company. I’ve worked with a couple of companies where they were growing so fast that, you know, there were points in time where the original optimized pricing and optimized ordering set got blown out of the water, you know, within six to nine months of it being set. And when that comes into play, you know, those are things that have to be both communicated to the supplier and that comes down to, you know, what are you going to do from an increased capacity perspective. But then you also have to relay that back to, you know, your upstream channels and say, here’s what I can do and this is what I can support in certain timeframes. And then that allows them to work with what they have until that increased capacity or that increased support for what they’re ordering comes online.

Will:
Yeah, if you, for a lot of, Yeah, go ahead.

Dan:
Go ahead. I was gonna say, if, you know, if marketing knows that there’s gonna be a big like PR blitz or a campaign of some sort, you gotta tell the supply team because you, you,

Will:
Yeah, that’s,

Dan:
I mean it sure, unexpected things can arise and of course, but the one thing you don’t want is to run app stock any point. So what suppliers really want our quarterly forecasts? I mean, half a year is more preferred, but first, in order to do that you got you extra have to understand what your, what your lead times are before anything. That’s, that’s the main thing. Cause you’re looking, so Ford out marketing is more focused on, on the now supply is, I mean I was looking nine months out and allow forecasts. So like every, once Q1 was over, you’re already into some holiday plannings already, which is, could be a lot to think about, but you’re just so forward thinking in your planning just because your lead times can take, I mean I lead times that took over six months at times. So you have to be thinking so far advanced in communicating with your suppliers on what, on what’s could happen now. Sure you could over order that does happen, but that offsets over time. It shouldn’t be hopefully that big of a cash detriment, but supply forecast is so forward thinking. Demand forecast is sometimes right now, oh my God, we had a bad month, which ratcheted up or we had a great month, let’s sustain it. But that doesn’t help us supply site cuz the, the months don’t overlap usually.

Will:
Yeah, I think if there’s something that you’re taking away from here, it’s that, you know, your marketing teams and your operation teams are only gonna be benefited by communication and letting each other know where they stand and what’s needed in the future. If at any point in time you decide, you know, you’re going dark or you’re gonna work on this and then share it once you’ve landed it, there’s a great chance that A, you may not be able to fulfill that order, or B, you may have a capacity issue to where you have too much and you’ve outlied resources or it’s gonna be something that you can’t support. I think Dan, that’s a good way to segment into you know, if you’re looking to get visibility related to some of those orders and then understand actually how much resources you have committed to certain groups. Avil does a great job where you can go look within all your purchase orders, you can actually then see how many of certain units you have on order and then communicate that easily back to your team, run reports, whatever have you. And then get some great visibility there. Dan, if people wanna reach out to us, where can they do?

Dan:
So let me get, let me pull up your cell phone number, like let them know, but no, seriously, you can reach out to us at hello@anvyl.com. That’s A-N-V-Y-L.com. And we’ll get back to you as soon as possible based off of your inquiry.

Will:
There you go. All right, well we look forward to hearing from you and to the next episode. See you later.