Latest / Inside Commerce: Ecommerce Strategy, CX and Technology Podcast / EP174: How Hypersonix Pricing and Inventory AI Enables Ecommerce Profit Optimization
Transcript
- 0:02Funds for lesson 3, platform podcast, sponsored by traction,
- 0:05hypersonics, and hosted by myself.
- 0:06James gerd, warm, welcome. If it's your first episode, fags
- 0:09are regular listeners. That we put a lot of time and
- 0:12effort into bringing this amazing content and Brilliant
- 0:14guests for completely free of charge.
- 0:16So please help us keep the podcast.
- 0:18Go in refers to other people in your network.
- 0:20You know, spread the word, get more people listening and help
- 0:23us by sharing or post will be put on social media, helps
- 0:26amplify. And also, please do write the
- 0:28show on Apple Spotify, YouTube. Or wherever else you're
- 0:30listening that really does boost your visibility with that mind.
- 0:33We hope you enjoy our latest episode.
- 0:35So our topic today is around like pricing inventory AI for
- 0:39profit optimization. So quite an interesting topic
- 0:41and let's face it, we all know that trading is tough in a lot
- 0:44of markets right now. Some sectors specifically, like,
- 0:47housewares really impacted by the prevailing economic
- 0:50situation, therefore focus on profit is logical and necessary.
- 0:54I mean should always be, but we know that in economic times like
- 0:57this. There's a sharper Focus.
- 0:59So today we're talking To our sponsor hypersonics, who provide
- 1:01profit optimization technology solution for econ businesses, to
- 1:04really couple to find out more about how it works and how
- 1:07people are using it. So we're going to cover how they
- 1:09Noble Prophet optimization using AI, am L to identify potential
- 1:13business risks. You need to address an
- 1:15integration parcel, put a road map.
- 1:17So let's welcome our I'll guess so first, let's welcome.
- 1:20Brenda is a co-founder and CEO and also I believe you're a
- 1:22council member of the Forbes Technology Council.
- 1:25So welcome from. Thanks Jim.
- 1:27Thanks for having us. It's great to be here, and it's
- 1:31great to be on this on this forum.
- 1:34Yeah, I've been with e-commerce and and enterprise software for
- 1:39more than a couple of decades. Now, been in the entrepreneurial
- 1:42world for over 15 years. I'm excited to be on the show
- 1:46and excited. Talk about what we can do for
- 1:49helping almost Merchants. Thank you.
- 1:51Yeah, I really recruited. Thanks for taking some choice
- 1:53and also, welcome to Rob who csdb and general manager.
- 1:58I also notice Rob, when I was two it still.
- 2:00Can you that you're a keen rugby fan and a rugby coach?
- 2:03Yeah, I am. I've been bit by the rugby bug.
- 2:07Probably 40 years again. Yeah.
- 2:09And no longer play, but, you know, my boys had and that I
- 2:15coach them all the way through their senior years in high
- 2:19school and down still. Do something about rugby every
- 2:23day. So yeah, that's a One.
- 2:28Heck of again. Yeah, find a legend or we use up
- 2:31all over. It's at yeah.
- 2:33That's got. Yeah, exactly.
- 2:34Been up. Half an hour room rugby tactics
- 2:36of rules before we get into too much product and Tony.
- 2:41Jordan, give us Slide the the summary for people who haven't
- 2:43come across either songs before. What what, what are, what is it?
- 2:46What do you do? Yeah.
- 2:52So James been, we I'll tell you the story.
- 2:57I am over this whole company started and why we are relevant
- 3:00today. You know, when we started this
- 3:02business, the idea was very simply to look at what what
- 3:06companies like you know, Amazon or Walmart or Target do for
- 3:10themselves and make make that democratized for every other
- 3:12e-commerce business out there. And and what that simply meant
- 3:16was, you know, you have millions of merchants right now.
- 3:19Trying to sell stuff online and a lot of times there are a lot
- 3:23of variables and signals in information right on what people
- 3:27are selling, how often are they selling and applying
- 3:31intelligence to that applying intelligence to understanding
- 3:34these signals and using that intelligence to turn that into
- 3:39recommendations that actually help them make that increased
- 3:42profit right. But you looking at Cost signals
- 3:45you're looking at you know selling signals you're looking
- 3:48at people signals Sylvia Supply chain signals you looking at all
- 3:51these various signals that are happening globally today.
- 3:54And how do you turn those signals into meaningful
- 3:57outcomes? Was the broad brush problem.
- 4:01We decided to solve and obviously, we are applied, what
- 4:05idea of artificial intelligence machine learning capabilities to
- 4:08be able to do that precisely accurately and in a timely
- 4:11fashion. But that's kind of what I
- 4:13personick does for e-commerce. And most importantly it's
- 4:16affordable, right? I mean, it's, um, it's one thing
- 4:19for a someone like a Walmart or a gap to have 10,000 people
- 4:25sitting in house, doing this, for them, right on a regular
- 4:28basis. It's another thing for a small
- 4:30Merchant or a midsize company to be able to do that spending a
- 4:36small fraction of that cost. And I think that's what again Ai
- 4:40and machine learning brings to the table is affordability at
- 4:42scale. And that's what we wanted to do
- 4:44right now. I think it is huge.
- 4:45Is mushrooming of e-commerce. Merchants nobly right, that's
- 4:49happening on Shopify. Platform on Bigcommerce on
- 4:52Magento on who covers and then all trying to make a living,
- 4:57they're all trying to be profitable, they're all trying
- 4:59to grow and we want to be part of that mission, right?
- 5:01We want to be part of the mission, enabling them to be
- 5:03successful. That's, that's our mission,
- 5:06right? And if you know, we are Michelle
- 5:08lines are their mission. I'm sure we'd all do well.
- 5:10And yeah, that's that's what the company is about.
- 5:14Bowl outside. Yeah I did say it's an interest
- 5:17in Focus because a lot of technology in e-commerce is
- 5:20capability-based rather than engineering decision making
- 5:24towards profit specifically. So can you I guess Robert be
- 5:27interesting for you to talk us through your obviously one of
- 5:30the different different sets of businesses.
- 5:33What type of business typically works with you and what size do
- 5:37you need to be X? Prime you just mentioned about
- 5:39it's not just about, you know, they're huge Walmart's of this
- 5:43and I think a lot of people probably think profit
- 5:45optimization sounds a bit bit out of my league.
- 5:47If I'm a small business so you have your talk us through like
- 5:50you, you know, cut your customer base types of businesses, you
- 5:53work with. Excellent.
- 5:54So we're, you know, June is brewed focused on product with
- 5:59when we get in invested into this business and really
- 6:05supported by our partners at by stinkin salting group and Intel
- 6:10Capital completely behind it because they're so, Vested in
- 6:13e-commerce space that we're working with companies that, you
- 6:17know, arranging in revenue from 1 million to 500 million dollars
- 6:21in a GMP. And typically from more
- 6:26precisely from AI, Nan ml perspective, we're able to find
- 6:33patterns in that data with as low as ten thousand transactions
- 6:38per year and 200 skews typically, When we're doing an
- 6:43implementation, we're starting out with at least two years
- 6:46worth of data. So our customers are able to get
- 6:49six to eight years with the data, which you know, AI loves
- 6:53the data, the more the better, right?
- 6:55And so when we begin our implementation downloaded the
- 7:00data, that we need from their e-commerce platforms, literally
- 7:04in under a day and then, based on the volume, our optimization,
- 7:08and Jim's start working, and within a week or so.
- 7:13We're coming out with something called recommendations and these
- 7:16recommendations are all profit based recommendations that are
- 7:20super easy for the the merchants to use.
- 7:24You know, they have five components to them that the AI
- 7:26engines put together and get everybody on the same page with
- 7:31regards to what the objective is and what the other reward is.
- 7:36And the coolest thing about it is that their perspective and
- 7:40their predictive, they're not Saying AI to looks at history
- 7:45and says, gosh James, you know, we're done running the numbers
- 7:49and we're really sorry to see what happened.
- 7:51Last quarter. I bet you were earlier here.
- 7:53Here, here and here. Yep, we're using historic data,
- 7:57but we're projecting and into the future of things that can be
- 8:01done. to help generate more profit every day for these
- 8:07e-tailers Yeah, I think well I think question this and maybe
- 8:12from this phone for you is because people I think a lot of
- 8:15businesses are used to that business intelligence BI
- 8:18Solutions whether it's part of an Erp or they, they bought a
- 8:21separate be like a power bi. And they're used to having to do
- 8:25analysis and trying to push this data to find where there are
- 8:29issues of opportunities Trends, Etc.
- 8:32So can you talk us through? How what's your profit is are
- 8:35doing differently to a traditional bi in terms of
- 8:38Intelligence space. No, it's a great question and I
- 8:41think that it helps clear the air on lot of confusion.
- 8:45Sometimes people have which we need Ai and bi but it also goes
- 8:49back to the fundamental difference of artificial
- 8:51intelligence versus business intelligence.
- 8:54So one of the just just in terms of this ground fast, right, I
- 8:57think what and artificial intelligence machine Learning
- 9:01System really does is precisely like the way it sounds.
- 9:04It actually brings in intelligence that it Ines
- 9:08automatically from data, right? So it looks at data and able to
- 9:13identify things like pattern since able to identify things
- 9:16like anomalies, it signed able to identify things like
- 9:19correlations and if then that then that right it's able to if
- 9:24then else. So it's able to do things that
- 9:28we're we're generally done by humans in the past.
- 9:31So if you use a bi system, it's a human being sitting and
- 9:35looking at the data and saying, oh, because I see this and I see
- 9:38this. Probably they are correlated or
- 9:40maybe I seen this trend chart. So first thing is you today, did
- 9:45you have a bi tool? It's a human being making these
- 9:47inferences. What a NES system does.
- 9:50Is it does that influence in for you, right?
- 9:53It's already like, Ruben all these patterns and where that
- 9:55becomes extremely powerful, is what I call identifying those
- 9:59needles. In a haystack, there are so many
- 10:01such needles in a haystack that even human eye cannot catch when
- 10:04you have so many volumes, the the level of volume of scale
- 10:07that we're talking about. Right?
- 10:08So many transactions, so many skus, certain sales velocities
- 10:12are dependent and patterns, you look at all these various
- 10:15signals and if using a bi to one of the biggest challenges are
- 10:18still, depending on a human being to look at the data and
- 10:21tell you what it works going on NES system, does that
- 10:23automatically. So that's Point.
- 10:25Number one. Point number two, that an AI
- 10:28system does. Like ours does is be predictive
- 10:31and prescriptive in nature. It's not just identifying those
- 10:35buttons and things. What a bi system does is
- 10:37generally looking at the rearview mirror.
- 10:39It's like driving a car, only looking at your rear-view
- 10:41mirror, right. That's that's not that's not the
- 10:43best way to drive a car. But I do you need to be able to
- 10:45see what's lying ahead of you. And that's what a system, like
- 10:49ours is not just providing the rearview mirror on things.
- 10:52We are looking at what, what lays ahead of you and that's
- 10:56where we're able to predict. And prescribe based on what
- 10:59we're seeing is happening, right?
- 11:01So, being able to look at all the signals, not just historical
- 11:04but also what's happening right now on your website?
- 11:07What are people clicking In one of the adding what are they
- 11:09abandoning looking at that. Matting that back with your
- 11:12forecast, Manning it back with their inventory, marrying it
- 11:16with your sales velocity and erastus Ethiopia items, tying,
- 11:19it all together and then presenting a view of the future.
- 11:23When you look through your front windshield you're seeing this is
- 11:25how it's going and this is what we prescribe right here in your
- 11:29data appear in Unit 8. The right here is what we will
- 11:32tell you to maximize that profit, right?
- 11:34That's the biggest difference in a nutshell.
- 11:36Those are the two big areas where Ai.
- 11:39And a system like ours is very different from a regular
- 11:43analytics or be Isis. Right?
- 11:46Attract, you'll Complete product Discovery growth engine create
- 11:51relevant shopping experiences that convert into sales and grow
- 11:54online Revenue with personalized, search
- 11:57merchandising, and recommendation Solutions powered
- 11:59by a I find out more at a tract.com Yeah, I think that's a
- 12:05nice differentiate will be released.
- 12:07You can you give a like a specific context around the
- 12:10stock bit, like an example of the type of intelligence, it
- 12:13gives people and is it then given a recommendation or is it
- 12:17more an alert for the person to act on?
- 12:21Yeah, I'm happy big to say, but I do that rather jump in as
- 12:24well. A good example is.
- 12:27So our system, for example, really has 11 different EA
- 12:29engines that work in cohesion, they all work together in making
- 12:34a recommendation. A good example of a
- 12:35recommendation on a prostitute from that.
- 12:37We would do is to say here are 15 items where we suggest you
- 12:41increase your price by 27% or white 11% and we do lutely add
- 12:47an item level. We would actually tell you these
- 12:49are 27 items that we You increase the price well, over
- 12:52the next 48 hours, and this is how much you should increase it
- 12:55by, right? And this is the forecasted
- 12:58impact and the forecasted lift that you're going to gain if you
- 13:01did that, right? So that's an example, of a
- 13:03recommendation we would do and you can actually click on it and
- 13:06say tell me why are you suggesting this?
- 13:09And we actually show you the why behind it.
- 13:11But I will tell you things like hey there's been certainly a
- 13:15sudden spike in traffic on your website, we suddenly see your
- 13:17sales velocity going off the roof against your Forecast, we
- 13:22are seeing their inventory levels are low for the side of
- 13:24me. See that your priced at less
- 13:25than your competition. So we'll give you all the
- 13:27reasons why we are making a recommendation.
- 13:31But being able to do this automatically, where the system
- 13:34comes to you and makes a recommendation, right?
- 13:36And it tells you exactly what you should be doing, and then
- 13:39it's up to you to make the decision, right?
- 13:41You can say yes, this makes perfect sense.
- 13:43I'm taking this decision. I, you know, you pick the ones
- 13:45you want and execute on and you can eat the price changes on
- 13:49your website in less than 30 seconds.
- 13:51So we're able to do that in an end-to-end fashion, where you
- 13:54can make that you see the recommendations, you make the
- 13:58judgment and you push it to your website.
- 14:00All this can happen within a matter of 10 to 15 minutes.
- 14:04That's kind of why we are here because of the merchants because
- 14:06Merchants don't have time, right?
- 14:08They don't have time. They don't have your they're
- 14:10trying to run a business and they want a system that can
- 14:12augment their decisioning and that's what we try to do, right?
- 14:19It's just the comment on that if I may, you know, we're
- 14:23typically, you know, in this economic situation right now,
- 14:27dealing with a lot of CEOs and we're working with the CEOs in
- 14:33that, you know, their companies are making a transformation from
- 14:36being really just excellent at building the top of the funnel
- 14:40and growing the company. And this transition is kind of
- 14:49Big change within their company to build this culture of
- 14:53profitability. And we've heard this from, I
- 14:56don't know crap. I mean, 120 CEO's that you know,
- 15:01we need to be able to, you know, start talking about
- 15:03profitability and make it a culture and it can't be
- 15:08something that happens on a monthly or quarterly basis.
- 15:10It needs to be an everyday thing.
- 15:13And, and so these recommendations that we make, we
- 15:17don't provide 10,000 a day. Day we provide for 25, but the
- 15:24ones that we do. How over a 90% probability or
- 15:31predictability score associated with it, using statistical,
- 15:35significance modeling, and in James to get everybody on the
- 15:40same page, it's everything the pain was just talking about
- 15:44these. These recommendations have five
- 15:46components to it and they form a sentence and it always has
- 15:50something to do with a product what products or categories of
- 15:54products. Is that the object number two
- 15:58is, there's needs to be an action that is made, and maybe
- 16:00it's market up, maybe it's marked down, maybe it's an
- 16:04inventory transfer, you know, maybe it's a, it's a purchase,
- 16:09you know, and then there's there's going to be profit that
- 16:12is going to be gained and we literally will go out in the
- 16:16day. I engines will Introduced a
- 16:20predicted profit number that will be achieved through this
- 16:25recommendation. And then it also gives you
- 16:27duration if this is say that you're going to do it sometime
- 16:30within the next 10 years. It will, it will name the time,
- 16:34you know, if you do this recommendation and it will come
- 16:37true in three and a half weeks from, right?
- 16:40And then the last is its established confidence score
- 16:46associated with that. And that in itself is helping
- 16:50our customers, all get on the same page with what the business
- 16:54objective is. And that is the heart and kind
- 16:57of the hero of our system is starting at this at this point
- 17:02and making it very clear, right? Yeah.
- 17:05Mike's not makes perfect sense and you will see talked about
- 17:09Native historical data because you'll know need to choo-choo
- 17:12train up. Our them, they need to have the
- 17:13data to make the decisions. What what is your?
- 17:17What's the standard? It's a black.
- 17:18What data do you need Awards? As a date, as a business, need
- 17:21to provide to you to make this actually work, make it being
- 17:25able to stop making recommendation if we're able to
- 17:28get, you know, soon as it multiple years.
- 17:31But we're always looking for transactions that went down that
- 17:35day. Any changes in inventory
- 17:39anything to do with the customer, either through loyalty
- 17:42or through US, CDP any changes to the product catalog.
- 17:47Anything in the By changing anything like on hand inbound or
- 17:52International Supply Chain. What more likely will always
- 17:56take clickstream data based on either, the behavioral Suburbia
- 18:03build a hypothetical behavioral demand forecast.
- 18:08We're all about profits so cost of goods sold but probably more
- 18:12importantly we're looking at true order valued and costs so
- 18:17things Like what the customer acquisition cost is, what is
- 18:21your cost of a partial shipment? Return, great refurbishment
- 18:24cost, restocking availability, wastage acts stored credit, and
- 18:29so forth that those are data that we're looking for.
- 18:32We also are looking for competitor data and we go out
- 18:36and actually provide that service has won out the modules
- 18:40that we have, and we'll go out and do the.
- 18:42The scraping that needs to be done looking for prices and
- 18:45price changes as well as inventory levels.
- 18:48Al's as we go through. So when you look at all of those
- 18:52elements, you know we're looking at in a nine different pieces of
- 18:55data that is brought into our API infrastructure in those 11
- 19:01ADI engines. Go to work on that data and
- 19:04daily produce new fresh relevant, highly predictable
- 19:09recommendations. And that does is some questions.
- 19:13I jab interested, do you, do you need somebody in the client side
- 19:18to be an expiration in data in order to drive?
- 19:21Or do you provide that service where you would do the data
- 19:24transformation to ingest it in whatever format you need?
- 19:30We are out there are no called implementation.
- 19:33So typically, so what we have done, we realized early on that
- 19:37are the two things that are Merchants.
- 19:39Don't have is time and a lot of resources to help get themselves
- 19:44starter, right? So what we did is we created, we
- 19:47have certified partnership, but all in Commerce Merchants, all
- 19:50e-commerce platforms. So we work seamlessly with
- 19:53Shopify work seamlessly with Magento with Bigcommerce.
- 19:57Woo, Commerce are all of them, right?
- 19:59We have. So we have a plug-in plate
- 20:01mechanism to work with all the data sources really, right?
- 20:05So we are able to do one click connects but most of them so we
- 20:09can pretty much get most of most Merchants stood up within a week
- 20:14Max met. You know, they have qumran
- 20:17systems problems. Going to be one a couple of more
- 20:19weeks for but in all cases there is zero effort from a customer
- 20:24standpoint or a merchant standpoint simply This will
- 20:28provide a plug-in API that connects to the source systems
- 20:31and pulls on the day of Ed, right.
- 20:34And we try to do that in a very quick time frame.
- 20:37So that way it's, it's you're able to see value within a
- 20:41within the first 30 days. Right.
- 20:43Our goal is always because you Merchants one, they're trying to
- 20:47big money every single day, right?
- 20:50At the end of the day you're trying to not lose a single day
- 20:53in taking the right decision. So how quickly can we get there,
- 20:56right? Is what we measure.
- 20:57Our ourselves on is to say we can quickly connect we can, and
- 21:01then all of our data signs engines well automatically as
- 21:04soon as we connect the data. So it's it runs every every
- 21:08night, and then it presents information that very next day.
- 21:12Come in everything from basic insights about a product to our
- 21:17recommendations, right? So typical customer today for
- 21:20us, get start seeing value within the first 30 days.
- 21:23And then kind of in start executing on those changes on an
- 21:27ongoing. Basis.
- 21:28Yeah, I know you referenced earlier, I know you talked a lot
- 21:32about on your website about Engineering Process to identify
- 21:36and flak risks specific. Can you give a few examples of
- 21:41what type of risk is it looking for like how do you define risk,
- 21:43basically and how this is working, how does identify and
- 21:47then what happens next to you other automation Strat, she's
- 21:51reacting, or is it more about alerting people to make them
- 21:54then make a decision? I'm connected.
- 21:58You shot at a real easy word and then you could, you could go to
- 22:01the top with separately, okay? Hit it.
- 22:05So, you know, here's a good one to understand what we're doing
- 22:09James. So when that data is imported in
- 22:12and into a AI engines every day, those is pregnant mentioned,
- 22:17we're that Ki engines are looking for anomalies in that
- 22:21data and over time, you know, they start, you know, Paying
- 22:26attention those anomalies and when they occur, you know at a
- 22:30high frequency chained together. It we hit it hits a point of
- 22:35statistical significance and it wakes up and it says something
- 22:38like hey the cell through velocity on product.
- 22:44XY and z are 12 percent above the demand forecasts and they
- 22:51have been for the last two and a half weeks.
- 22:53But we're letting you know, right now we Was it, you know,
- 22:56nothing's going to change. If you we all things the same.
- 23:00It's going to continue to do this and and the rest of the
- 23:02engines, say, all that is great. We love to see our product
- 23:06selling well, but the inventory, AI engine, keeping it simple
- 23:11might say, it's good news. However, if we keep selling it
- 23:15at this rate, we're going to have a stock out in three and a
- 23:19half weeks from now, we were planning on the replenishment
- 23:23border Being received within five weeks from now we need to
- 23:28get together and create a recommendation for James and
- 23:34that recommendation and is going to be something that is
- 23:38calculated as having the highest probability.
- 23:42But also the highest profitability will be presented
- 23:45to you because you can imagine they're probably about 10
- 23:48different scenarios that you can have there.
- 23:50And you know, going with you know what Mentioned before of
- 23:55that recommendation could be. We recommend that you increase
- 23:58your prices on these 12 products.
- 24:01By 7% for a twenty. One thousand dollar profit gain
- 24:07in the next 17 days with a 93 percent confidence score.
- 24:12Right? It could have went the other way
- 24:14who said, hey these prices for these, or these products price,
- 24:20elasticity is very high. Meaning that if you move your
- 24:23price, You know, even one percent, you know up, you may
- 24:28stall that product. So we need to make an inventory
- 24:31group and it couldn't be, you know, we have agreements with
- 24:35our distributors or manufacturers that expedite
- 24:38product in and though we're going to have to pay a fee.
- 24:42We will be able to get at least a quarter of our order in fast
- 24:47enough by paying that unprofitable see.
- 24:51But it's going to, you know, prevent kvi from selling out and
- 24:56having an outer stock situation. So it weighs the options and
- 25:01comes up with predictability and the profitability leader of
- 25:06those recommendations. That makes sense.
- 25:11Yeah, that's a really nice. Nice example, actually, because
- 25:15I was saying this from, from a lot of e-commerce businesses,
- 25:18but if you Inventory management is a constant challenge because
- 25:22you don't want too much because it's like it's lost its lost
- 25:27opportunity but you really don't want to to have fractured stock
- 25:31because then I start seeing your conversion rate.
- 25:33So you're nice example. And so prayer, many anything,
- 25:36any other examples? Yeah, there you can share the
- 25:39will help. People understand what are the
- 25:41risks that can be identified? Yeah, about and inventories,
- 25:45obviously, one area are where it's either.
- 25:47You know, we have an artist top prong or other opposite of it to
- 25:51brats point where your mixes inventory to right.
- 25:54I'll give you three up Sarpy working after all, I'll because
- 25:57you have locked, your working capital and up in your turn that
- 26:00into actual cash flow is an example of one of the other
- 26:04common things we see are two other areas that we have been
- 26:06seeing that become really popular is one is Understanding
- 26:11how your competitors the way and how's that influencing your
- 26:15sales, right? So one of the things that we try
- 26:17to do with our computer module is to monitor your top
- 26:21competitors. Look at what price changes they
- 26:24have made, or what? Hi, look at.
- 26:26If they are running a promotion or not, look at all these
- 26:29signals around yet on Twitter. At seeing that the influence of
- 26:32that on your sales, right? Is that impacting your sales?
- 26:36Do you need to act on your price a or your Promotion or whatever
- 26:41you're trying to do. So bringing in computer data and
- 26:45tying that into price promotion is again, an area where we try
- 26:50to see that if there are risks that are arising from that, that
- 26:53it back your sales, we try to highlight that.
- 26:56Right saying, hey, we see that there has been changes, what
- 26:59your competitors are doing and what you can do in a system like
- 27:02ours is you can, you can automatically set up what we
- 27:04call the business rules or thresholds where you could say,
- 27:07hey take this automatic action. I think about, you know, what
- 27:11example is think about clients? We have clients who have
- 27:14thousands of skills, right product at 57 10,000 schemes
- 27:19that they were dealing with and you move.
- 27:21It's impossible to sit at monitor every single skill.
- 27:24It gets even five other competitors, right?
- 27:27Be so you're basically looking at a combination of thousand
- 27:30plus Q, is I get spy veteran competitors on a daily basis or
- 27:33a weekly basis making price changes are making emotional
- 27:37changes, you know, that's just Humanly not possible, right?
- 27:41So you need a system that can actually monitor and Betty that
- 27:44these changes that are happening.
- 27:45And this thing act or this is a person to your bill overall, REM
- 27:48either might impact. So that's again, an area where
- 27:52we see a lot of activity around right.
- 27:56There are other risks. Also we try to kind of
- 27:58highlight, where there are we see items where suddenly sales
- 28:01have dropped to zero, we call it zero zero sales time and we look
- 28:06at what is the average demanded well between a certain item,
- 28:09right? What If an item has been
- 28:10constantly selling at a certain demand pattern and your average
- 28:14demand interval between selling, let's say the span has been
- 28:18always every seven days. You sold this one, one of the
- 28:21span and now all of a sudden you have installed something for 60
- 28:24to 90 days, right? Yeah.
- 28:27You are now stand, then we correlate that with inventory
- 28:30and we carry. Yeah.
- 28:31Also something spend sitting around and all of a sudden, the
- 28:33sales dropped to zero first Alaskan state, a right?
- 28:36So we need, we will highlight that and we take tell you What
- 28:40recommended actions we suggest for you to bring that back into
- 28:43the sales motion. So things like that, right?
- 28:46It being able to help me make this micro decisions.
- 28:49There are hundreds of micro decisions that need to the
- 28:51prophet and the the prophet goal and those are all the areas that
- 28:55we constantly bubble out. So you're always staying abreast
- 28:59or what decisions you need to make to kind of move the needle.
- 29:02Hypersonics helps e-commerce companies make more profit every
- 29:06day. This AI driven platform delivers
- 29:09recommendations for pricing and inventory that lead to bigger
- 29:13profits. Visit hypersonics, dot AI
- 29:16forward, slash podcast to get a free trial and have interest how
- 29:21to say, how do you make it a customer Centric?
- 29:24Because, I mean, that those are really nice examples, but
- 29:27there's always an anomaly to a particular customer.
- 29:31I mean this Say one example, a customers problems where they
- 29:33have a few that are just seasonal items are for, they
- 29:35have a you know, a standard sales pattern, then it drops off
- 29:39X months here. Then it comes back.
- 29:41How old is that intelligence get built in so they're not again
- 29:44like false Flags. Well, you know, we started out
- 29:50this company, early are in, just think about grocer, you know,
- 29:54when we were building these AI engines and stores, it's all
- 29:58about seasonality and the day I engines, it just had to be a
- 30:01part of the Or so, what we're looking at, especially if we're
- 30:05fortunate enough to get, you know, years worth of data, Rai
- 30:09engines are picking up, seasonality, demand curves, you
- 30:14know, throughout the year based on products, like products or
- 30:18categories of products, we also are influenced by our customers
- 30:22because they will give us their hypothetical dates for their
- 30:29seasonal items. And so we'll have to use That as
- 30:32a benchmark as well. But Rai is fully trained to to
- 30:36look at that. And obviously perishable goods
- 30:38are really short season. You know, avocados going to
- 30:41last, you know four to five days in the store, right?
- 30:45And so we're timing of a product were to season this kind of in a
- 30:50second nature to us and it was built in at the beginning of the
- 30:54architecture. They price moment ringside for
- 30:58the brand sudden in say Grocers is a massive ongoing daily
- 31:04challenge for them because the number of campaigns promotions
- 31:07competitive price drops. Yeah, we don't lines like that,
- 31:12right? And we start thinking a lark
- 31:14perishable goods are Goods that even have an expiry date, right?
- 31:17Like there are certain things that have expired, a certain
- 31:19things. I mean, the write the value of
- 31:21rocks 20 after that date, right? So so they have to make
- 31:25decisions. I'll correlate that with their
- 31:27inventory levels and make it appear institutions or Mom
- 31:31decisions and we help them with the world markdown planning
- 31:34process. So we haven't, we have module
- 31:36that helps serve with the ideas markdown strategy, right?
- 31:39What would be the? You can do a multi period mark
- 31:41down you can get clearance mall. Now we do all of that stuff,
- 31:44right? So again in all of the cases, we
- 31:46are trying to always say what's the most profitable route where
- 31:50YouTube. Dang.
- 31:51Right. Each week.
- 31:53Yeah. Yeah.
- 31:54Mate. And you did you do the title?
- 31:56Integrations. Get all of these and you said is
- 31:58ongoing signals. It's not just historic.
- 32:00Is that like real time so just like an analytics package,
- 32:03you'll pick it up the session in the browser.
- 32:05The so you'll sell day to go because I got a question and
- 32:08she's one of the one of the issues I've seen for a big
- 32:11businesses is where marketing and trading not always be
- 32:14talking the same same time. Which so a marketer is something
- 32:19launches really really good successful, paid media campaign,
- 32:22that smashes sales and items of the the trading And the bindings
- 32:27hadn't thought they were going to suddenly need to rapidly
- 32:29replenish it. Links back to your point
- 32:31earlier. Let's do it.
- 32:33I'm assuming but assumptions dangerous bird.
- 32:36The algorithms are looking at the underlying drivers of the
- 32:39purchase if they suddenly see a big spike in traffic from high
- 32:41converting, is it able to then change its prediction model and
- 32:46go? We think we could set up quicker
- 32:48now cause somebody's push the button.
- 32:50Yeah, yeah it is it that happens all the time and it's not
- 32:54intended but but things like looking at that, the page view
- 33:03of product detail pages and the AI engine picks that up and says
- 33:08cheese you know for these products were seeing at 28% view
- 33:14rate and actually there on the site you know 30% longer when
- 33:20they're looking at these items and it might not be something
- 33:24that you know the whole words of Isaac Understands because they
- 33:28haven't spoken to marketing, but we will highlight that
- 33:31heightened activity so that they're able to then look at it
- 33:35and say, oh okay, what what's going on here before?
- 33:38We take action on this. Let's see if there is a
- 33:41promotion that's been pumping up the view in this because we've
- 33:46kept, our inventory is level and we kept our price.
- 33:50The same, you know, something is driving.
- 33:53And it could be a situation where I don't know, baby.
- 33:56These products from seen on TV and they got really popular.
- 33:59All of a sudden maybe it wasn't a promotion and maybe our
- 34:02inventory levels of drop by 30% and therefore, we're probably
- 34:07going to make a price recommendation to maybe even
- 34:12increase, you know, the price in order to create more
- 34:16profitability and also have a an expedited, you know, purchase
- 34:21requisition to get more inventory in to cover that
- 34:24Demand right. But but being able to look at
- 34:27what that consumer is doing and understand a purchase intent, is
- 34:31an important part of our formula.
- 34:34Yeah, that makes that I think it's a nice point because what,
- 34:39when you're going towards profit there are time when Tarzan
- 34:42businesses can sell less but more profitably and that's good.
- 34:46There's other times where they want to drive acquisition
- 34:49because they're hedging their profit on future retention and
- 34:51monetization of lifetime value. So, yeah, it's interesting, the
- 34:56people, I guess this is the beauty of, of the AI am L was
- 35:00been able to do it at scale and present opportunities that you
- 35:03can. Then decide which strategy best
- 35:05fits with your current current goals now, that's exactly right.
- 35:09I mean, that's that's you hit it on the head, right?
- 35:11And that's the real value. That an AI system brings is the
- 35:16scale and being able to do that when you're dealing with so many
- 35:21students. So many different signals and so
- 35:23many transactions on a daily or monthly basis.
- 35:27And to be able to that affordably to write being is or,
- 35:31you know, it's, it's one thing to have a team of People sitting
- 35:34and doing it right? And still having errors in it
- 35:37and it's another thing to do it with in Split seconds.
- 35:40At a fraction of the cost and being able to make smarter
- 35:43decisions, right? So I think those are the areas
- 35:46that really is changing the game and we are seeing this, you
- 35:49know, ongoing growth in chat Jeeping tea and AI as a jest
- 35:54General Trend across the world, right?
- 35:56Various saying that, hey, AI is changing the way generative.
- 36:00AI is changing the way how people are consuming Zooming
- 36:04information making decisions. It's automating tasks and
- 36:07helping augment lot of people intelligence our and you try to
- 36:11do the same what much more quantitative data, right?
- 36:14As lot. More quantitative based we focus
- 36:16on things that are really driven underwriting profit or owners
- 36:20and Merchants and Merchandising leaders and we help augment
- 36:24their decision process, right? They clear, you know their
- 36:26business better than anyone else and our job is to augment that
- 36:30with the right information. Yeah I think I That's where you
- 36:34notice. I put in because AI there to
- 36:36help humans work smarter and it replaces some processes, but it
- 36:41doesn't take over the need that someone's got to have oversight
- 36:44because there are times you need to decide which of the
- 36:48recommendations to bullying which so kissable.
- 36:50We all know that sometimes selling out is a good thing and
- 36:52not having stopped going to think as you don't want to.
- 36:55Yeah, I mean they don't know we never know anyone's business.
- 36:58Like the owner of that business. Let's just give it a break.
- 37:01There's no way they'd give us a system can do.
- 37:03But what a system can help is to augment that decisioning you
- 37:08just be lot. More productive radio, just
- 37:10going to be a lot more productive, a lot more faster
- 37:12and you able to pick an opportunity at the right time,
- 37:15you haven't figured out because there's a window during which
- 37:17this evaporates the opportunity evaporates, right?
- 37:19It's kind of like stop. Picking you buy something today?
- 37:21It goes up tomorrow but if you didn't figure out today your
- 37:25kind of locks that opportunity, right?
- 37:26So there's a window of opportunity we're going to
- 37:29optimize and capitalize and that's what we help them
- 37:32achieve, right? I also like the fact that you
- 37:35can automate some of this because, for example, if you've
- 37:37got if you flag certain skews as clearance and you want to get
- 37:40rid of them and you know, that by reducing the price, you're
- 37:44going to drive up your sales that you can just automate that
- 37:48and not worry about it. Whereas your current lines, you
- 37:51don't want to be the value in by reducing the price too early in
- 37:54the season. So I think that's nice.
- 37:56Then let's, let's talk about the integration piece because this,
- 37:59I'm always always interested in as an e-commerce Consulting.
- 38:03Alton and U. Prime you mentioned earlier
- 38:05about some of the platforms you work with.
- 38:06So Rob, can you talk people through?
- 38:09What are the options for an e-commerce team to implement
- 38:12Hypersonic? So what what do they need to do?
- 38:15Is it all connect to base? Where they just literally
- 38:18connect the existing store to your product.
- 38:22What do they need to think about?
- 38:23In terms of inspiration, you have.
- 38:25So we're we spent our money Reba.
- 38:29Latrice quite a bit. Was obviously the a I was
- 38:33structure. Number two was ease-of-use, get
- 38:38everybody on the same page, understand what these
- 38:40recommendations are and how to accept them, you know,
- 38:44investigate them or reject them. And then number three was making
- 38:50sure that it was a no goat implementation.
- 38:52So we invested heavily in to, you know, building adapters and
- 38:58connections with Shopify. Bigcommerce will Commerce
- 39:02Magento, Salesforce Commerce, Cloud Channel advisor.
- 39:07You know, I don't know if I mentioned Magento yet Magento
- 39:11to. We looked at all, the primary
- 39:14players and built that, so things are happening quickly,
- 39:19what? Was part of our edict.
- 39:21When we did our voice of the customer and talk to CEOs, they
- 39:24just abhor large projects that run off track both in time and
- 39:31end in spam. So that's where we get a lot of
- 39:35that money. We're also, you know, just from
- 39:37an analytic standpoint, you know we always want to get Google tag
- 39:42manager for Google analytics or omniture mix.
- 39:44Bout, you know, any of those L Linux packages That we can grab
- 39:50data from and bring in or that we could, you know, export to
- 39:56definitely from a business managed systems and erps, you
- 39:59know, next week on like, Road soft, Dynamics QuickBooks, you
- 40:05know, integrating into those and actually just seen seven or
- 40:08cin7, excuse me, dear systems, just finished it, and it
- 40:15negation on a partnership with them as well.
- 40:17They got about 10,000 customers. That are out there as well.
- 40:21So we'll be looking at replenishment inventory at so
- 40:24forth from from them of their Erp.
- 40:26We also support you know, that the major Market places from
- 40:29Amazon Walmart, you know, working through Miracle.
- 40:33We know that there's strong player in that market as well.
- 40:36Are, you know, Wayfarer and then for, you know, I'm talking to a
- 40:40couple of companies that, you know, are using some pretty
- 40:44older Technologies, you know, one is Ruby on Rails, you know,
- 40:48and For example, you know, in those situations, you know, we
- 40:52could build and a custom API, using our API adapter, bringing
- 40:56it out that some of them might even use the CSV or we would do
- 41:01something in FTP. So we've made it really defined
- 41:05by by the package and the target system that were integrated, 22
- 41:11is soft and flexible. You know, tools that are there
- 41:15available in order to have successful implementations So we
- 41:20just don't want to waste a lot of time that and we have you
- 41:22know, excellent people that have been with us for years that
- 41:25understand the science of data and Urdu.
- 41:29No. Excellent at that.
- 41:30But we wanted to automate as much as we could.
- 41:34Yeah, I think I recall most Merchants will not say who they
- 41:38are and they are think gone are the days of the expectations.
- 41:40And the integration fees are more than the development fees
- 41:43and design phase. Yeah, my question years were all
- 41:47in PRP in that, that was even a generous, right?
- 41:52So every decade, which is now this is the end of my third
- 41:56decade in that technology. It's it.
- 42:01It's been shorter and shorter. So this was a huge goal of mine
- 42:03to A product that Growth Company to wear.
- 42:06Yeah, that was a non-issue were making it less loudly.
- 42:10Israel's 1X question Prime, which is product strategy
- 42:14roadmap, where it where you Takin Hypersonic?
- 42:18So I what can people expect is it more about yo you just
- 42:22focusing on key capabilities and Tool in and performance or is
- 42:27about enhancing feature sets. Yeah, I know it's a great
- 42:30question and obviously I think we're just scratching the
- 42:33surface. Surface on areas where a I can
- 42:36influence e-commerce. Where I think is it bring
- 42:39eCommerce as a market is growing in AI as technology is getting
- 42:43more and more widespread, right? And it's getting more and more
- 42:45widespread adoption to be rhombus leading the way when it
- 42:49comes to anything to do with sales and profitability, when it
- 42:53comes to e-commerce, you know, we are adding more and more.
- 42:57So, the points that you brought capabilities around
- 43:00understanding marketing attribution, then the Right
- 43:04understanding. Hey, because that's influence in
- 43:06profitability to Rob's earlier point, right?
- 43:09Customer acquisition cost cost of, you know, ads return on ad,
- 43:15spend and looking at all the. Because when I think about
- 43:17profitability, for e-commerce, Merchants obviously, this cost
- 43:20of goods, but then there's a cost of customer acquisition
- 43:23rate in these all these These are probably their biggest cost.
- 43:27And and so we're trying to always look at all aspects of
- 43:30profitability, right? And trying to add capabilities
- 43:33and within the product. Act that gives them this
- 43:35holistic view of, how should they do better selling in better
- 43:37marketing, right? Then make a mark in and better
- 43:40sell, they will make more profits because they understand
- 43:42the products that they're selling and that's our goal,
- 43:45right? Our goal is to add more
- 43:47capabilities in the product that gives them full picture of their
- 43:50profitability, understand, it helps them understand what to
- 43:54Market, how to Market, what to sell out to sell, right?
- 43:58And if we can help them do that in a cost-effective manner.
- 44:01In a quick time to Valley with no code.
- 44:04That's, that's our goal. And we're going to add product
- 44:06features towards that, that gives them that, you know, One,
- 44:10Stop Shop solution from us, that we cover quite a lot today.
- 44:16I know that it's hard to say, it's hard to simplify something
- 44:19that that's doing quite a lot of complex computations.
- 44:24But inevitably, there will be people who think more qualified
- 44:27to find out more. So if someone wants to see a
- 44:30demo so they can actually see what it's really doing their own
- 44:33eyes. How they get in touch through,
- 44:35they reach out to. Yeah, they just aren't going to
- 44:39work at Sea and you know what? It's we don't really show a lot
- 44:43of part sure, our product and people get it within just a few
- 44:48minutes. So it's it's fun to watch that
- 44:51reaction but where they would go, James is to hypersonics and
- 44:56it's with an X Out of not a Sikh Hypersonic dot Ai and then
- 45:01forward slash podcast and fill out.
- 45:04The form and we will get on it immediately and set up that that
- 45:10meeting is quickly as possible. So that's where we go now.
- 45:13So there's hopefully, you know, there's something bad Yeah, even
- 45:20going out to our site, you'll be able to do that.
- 45:22You'll see request a demo all over the cell, so pretty Thanks
- 45:27and I'm primarily. Thanks so much for taking the
- 45:30time for answering the questions of enjoyed it.
- 45:33I've learned learn new stuff, they but she's always giving us
- 45:35why I love doing these podcasts and as I as Rob said, anybody
- 45:40listening wants to find out more get demo.
- 45:43Go to Hypersonic stop II for trash podcast and we'll put a
- 45:47link down the landing page as well on our website and thanks
- 45:52as always everyone for listening.
- 45:53I hope you enjoyed it. Keep an ear out for next
- 45:55episode. We drop one every week and don't
- 45:56forget to give Us a traitor on Apple spot Ride YouTube or
- 45:59wherever your listing. Take care, everybody.
- 46:09More information on this topic, head over to re-platform dot.
- 46:12F m, for our audio podcasts to discuss a project or if you like
- 46:15to chat about any of the topics covered in this episode in more
- 46:18detail. Please reach out to myself,
- 46:19James guard. And my co-host put Rogers, via
- 46:22LinkedIn and Twitter. Thanks again for listening and
- 46:24keep your ears peeled for the next episode.