Host Mike Graen is joined by Barbara Chase - Vice President, Global Sales Store Intelligence at ReTech Labs, Inc./SymphonyAI - for the final part of their conversation about leveraging computer vision to identify OSA alerts for resolution including:

• Business Justifciation for Computer Vision (1:25)
• Data monetization for shelf product data. (5:30)
• Who is their customer? (9:12)
• How to get started. (15:21)

Greetings. My name is Mike Graen. Welcome to

another edition of the podcast focusing on on shelf

availability sponsored by the Walton Supply Chain Center at

the University of Arkansas. Today, we are excited to have

Barbara Chase with us. Barbara is a longtime professional in

the on shelf availability area within the industry. She's

currently Vice President of Global Sales for ReTech, or

Symphony AI. We talked to her about leveraging computer vision

in store to understand things like out of stocks, or incorrect

products, or pricing discrepancies, etc. Let's join a

conversation with Barbara. That, you know, when you start looking

at robots running around in stores and fixed cameras dropped

from the ceilings or people coming in and taking pictures, I

can't get enough people to actually stock the shelves now

how am I gonna get people to come and take pictures? And then

I think about this word that just sounds expensive:

artificial intelligence. I mean, it just sounds expensive. Right?

So how do you justify that? How do you do the business

justification to say this is a an investment that you need to

move forward?

Well, what I can also say is we have a value

assessment team that consults for free to help with those

business justifications. So we actually have a quite detailed

ROI Return on Investment assessment. So it's based on

based on annual revenues, it's based on the number of SKUs,

categories, frequency of images, and then based on the numbers

that we've proven with our, with our customers. So for, for

instance, hang on, I'm just gonna use my PowerPoint as a

guide here, but I don't need to share this picture is we have,

we have proven up to 5% in increased sales, and we which is

which is huge. And, and this is how we look at it as well. So

from that it's kind of a domino effect, where it's a, when you

have increased planogram and compliance, and we've got a

recorded up to 23% increase in that, then that correlates to an

increased on shelf availability, or OSA and we've proven up to

11% of that. And then that domino effect is in fact, your

sales. So you can take a percentage and it is it is

accepted and embraced by retailers and consultants a

percentage of that. If you take like a 1% OSA that will equate

to a certain percentage in sales. And so and again, there's

lots to back that up. That's not just Barbara's idea. And so I

will say that is a key, a key differentiator. And that's why

too there are so many pilots going on, because people want to

say in my environment, you know, in my categories or entire

store, whatever, you know, we have proven an increased OSA,

you know, worn increased planogram. We also from an

optimized labor, so, so to if we can say, hey, now that we have

these prioritized workless, we don't have to have or, or if

we're using shelf cameras or robots, we don't need the labor

to be doing that. Regular I'll just say regular whether it's

daily twice a day weekly, doesn't matter. But that regular

walk of as I mentioned earlier, shooting holes or gaps scanning

or whatever that goes away. And then the people that are also

going through checking prices, that goes away. So those are

hard, hard dollar revenue. Another -

Let me interrupt you real quick; because anytime you

say reducing labor people relate that to we're going to automate

people and get rid of people. I've never seen a retailer yet

get rid of people because of this. What they do is they give

them more meaningful tasks than doing doing audits which robots

and things can do. Because there's always more work to be

done in a retail store. You can possibly get that. So whenever

you say reduce labor - just to be clear for anybody listening

to this - that doesn't mean get rid of people. It gives them

more more empowering and hopefully customer-facing jobs.

Or helping more... well, you know, picking orders for

customers and things like that not just scanning a bunch of shelves.

Exactly. And I'm so glad you pointed that out is,

as I mentioned earlier, they may have 100 employees, but they

really do need 500. So that's where again, we want them on

customer-facing or making sure that they're restocking. Because

we don't, you know, we're, we're, we're, we're the brains,

but we're not the arms and the hands, you know that that's

where we need, we need that help. So it's it's definitely a

collaboration between the two. And that's why there's so much

excitement. Now, the other thing from a business case perspective

is, what we have that is unique is data monetization. So we have

just stood up a shelf product data monetization service. So

with Symphony AI, and I think the excitement about us being

acquired about a year and a half ago is, is we've just got this

army of people like the value assessment team for free to work

through those business justifications. But also,

there's a monetization team already working with and have

has coaches working with retailers and CPGs for loyalty

data. So now we are adding the shelf data. So we say store

intelligence, because we do displays and bunkers and all

that kind of stuff. But but just to help decipher between what's

typically known as data monetization, and shelf product,

data monetization, it's worthwhile. So when you've got

the like, the top 10, or 11, CPGs, that have, you know, 1500

... 1000 ... 500 sales reps, you know, these DSD's are just gonna

go do things on their own, you know, blah, blah, blah. But for

a tier two, and tier three, they don't have those sales forces.

So they want to, they want to buy that data. And oh, by the

way, even for the big guys, they still there might be certain

areas, whether it's a niche from a geography or a demographic,

that that data is still worthwhile to them. And rather

than having their own people or an auditor, or broker, well, if

I can just buy it, like I do from other data providers,

that's, that's valuable and worthwhile to me. One, because

as I showed that virtual visit, maybe I don't need to visit a

store quite as often. And I as a CPG, then can cut my travel

expenses, because now I'm getting a regular picture,

whether that's daily, weekly, bi monthly, monthly, whatever that

that frequency is. And so that's very exciting. That is new top

line revenue, and in a couple of our pilots, right now, you have

to do a pilot first to just to get the data to be able to

socialize it with with the different CPGs. But then having

that it's it's launching those conversations. So the idea is

not that that our solution costs anything at all. The the the

plan, is that it is a revenue generator. For the retailer, for

the retailer.

Okay. Got it. Excellent. Wow. Whew, there's a

lot of stuff you got going on. So your primary customer, I

guess there's a couple things, right: number one, it's the

retailer that wants this intelligence. But then you have

to obviously, collaborate very closely with the data collection

folks, whether it's the fixed camera folks or the robotic

companies, etc. Because I would assume for systemically

collecting that data every single day to every single

shelf, etcetera. Capturing with either a camera in the ceiling

or fixed camera on the shelf or robotic scan is probably the

probably the more preferred method. Would you think so?

Who's your customer? Who is your customer I guess the end of

this?

Yeah. And I'll tell you the, the, the answer is

all over the place. Because I said, you know, retailers are

unique. And here's what's interesting, too, is: do you get

trucks every day? Because if you don't get trucks every day, and

there's not to action, it might might not make sense to have a

daily picture. It might make sense to have a weekly picture.

Or maybe if your trucks come two or three times a week because

when we're talking about supply chain, right, let's you know,

follow that through. And so, so the, the you don't want data for

data's sake; you really need to understand the entire ecosystem.

And then what's the sensibility in that - you know - the entire

story. Because different categories have at much

different built in different velocities. And that's what's

kind of fun is you, you can, you know, you can look at your

stores and have that surgical approach in regard to, as we

said earlier, how you image capture then to be able to staff

to be able to act against that data. So that you are, you know,

always keeping customer satisfaction at the forefront.

Yep. Oh, that's awesome. That's awesome. So I

guess one of the questions that I like to ask you is, you know,

what, have what question Have I not asked you that I should

have? That's one of my favorite questions at the end. We got

about 10 minutes left here, I want to make sure. I've asked

you a lot of questions. It's been a conversation, which is

really the way we want this to be. This is Matt's dream to be

able to have not a bunch of PowerPoints but a conversation

about real retail, on shelf availability challenges. Is

there a question that I should have asked that I didn't?

I think it is, in fact around - so yes - and I

think it would be around 'define computer vision.' Because there

are many solutions out there, and they say computer vision and

AI, but they don't have the ability to, in fact, recognize

that product, they do a great job at recognizing, "hey,

there's a hole here" could be because of a misplaced item. But

hey, if there is a hole there, and let's - and and that's good,

that's goodness - but you can do better by understanding what

that product is, and where you know, better management

organization of the store, by having true - what we call true

computer vision. It just adds to the use cases, because I think

that's also the difference, you know, a lot of time as I say, I

feel like a grandma in this space, because I've been in it

for five plus years, and the speed at which it moves. That is

a super long time. And, and just even you know, from five years

ago, compared to today, the ability of AI, or even just; we

were in a retailer, a couple of weeks ago, and the aisles were

really narrow. Well, you know, in 2020, which is, which is

relatively recently, they introduced negative zoom in

regard to tablets and cameras. So before negative zoom, we

wouldn't have been able to help that retailer. But now it's,

it's no problem. And so so with that, it's that understanding of

how as a as a retailer, or CPG, if I am going to explore and,

and potentially invest in this, I want to maximize the use

cases. And the only way to do that is to have that combination

of AI (that's good, prioritize workloads, volumes of data to

make sense of it), true computer vision (I not only recognize a

whole but I recognize the product Honey Nut Cheerios to

Cheerios and where it goes), and optical character recognition

(pricing promotions, that's big). To your point, as you

mentioned earlier, 'hey, I came in it's Wednesday and you know,

where is it? And then the other question- if I can say two maybe

missed - is how do I get started?

Ok. Barbara, how do I get started?

Well ... what myself and my team would

recommend is a pilot and what I what I say is is we recommend a

Crawl-Walk-Run approach. And as fast as AI is happening and with

the kind of chaos that snap or not snap ChatGPT. My... my

daughter showed me that there was a Snapchat and ChatGPT

that's what was on the top of my head just like a week ago and I

was like what? So you know, so now that that's caused like a

flurry of, okay, I understand now what that means to me

personally. So now I want to understand that: pilot. How does

it apply to your specific brand and your specific ... so that

then it's a you know, it's not only a meet-and-greet. It's a -

okay, let's really understand your operations, your staffing,

what systems you have, and how then does that work together in

the entire supply chain? Because now this is the last piece like

I mentioned earlier, it's never previously been available

before. So supply chain has just been blind. They've just been

blind to the most important piece: me as a customer, what do

I see? Right? So now that that's there, then we can understand

and a pilot, which takes about three months, that from start to

finish, how this works, and then two: how to make a defendable

business justification for a rollout.

There's a word that you didn't say, and I was really

happy about it. You're ready for it? Oh several words: POC -

proof of concept. What you didn't say that. And here's why

I think he didn't say that. We don't need to prove that this

works anymore. Somebody asked me to do a proof of concept with an

RFID project. We've been doing this for four years, we know it

works. The only question is, can we set up the right parameters

in a pilot to figure out, you know, how do we make this work

for you as a retailer. So thank you for not saying we'll do a

POC, what you said was we'll do a pilot, I think the

Crawl-Walk-Run analogy is a really good one in this kind of

space.

Yeah. And I think it just makes every we don't,

we're so confident, we don't have any hooks in it. If it

doesn't work out for whatever reason, we just take our toys

and and go home. So there's just really no stress, it's all

excitement, and then just to even be involved, from an

employee perspective, whether it's in the store, or whether

it's up the headquarters chain, oh my gosh, like everybody's

like, do I do I get this on my resume? Do I get this

experience? And it should be that exciting, because there is

so much to glean, and to learn and how cool if your

organization, you know, supports this exploration, to understand

how to continue to be fiscally responsible to make your chain

as profitable as it can be starting with customer

satisfaction. And so yeah, we've had a yeah, it's, I don't sleep

much at night, because I started the day with London, Germany and

France. And I end the day with Beijing and Japan and Australia.

So it's, it's an exciting time to be here. And I just would

love the opportunity to talk to as many as possible. So they can

- I always say seeing is believing - so that we can, we

can create more believers.

Yeah, this is awesome. Well, let me summarize

this, because I had the good fortune of actually meeting and

working with Sam Walton while he was still alive. And one of the

things he said - and this quote has been shared a lot of times

but - "customer is ultimately the boss. They can choose to

hire us or fire us every day, by choosing to spend their money

with us or somewhere else." And number one factor I've heard of

people walking away from a retailer is front end checkout

experience takes too long, etc. And number two one is

availability of product.Didn't find the product. I was coming

in to look for this and after the third time of it not being

there to heck with it, I'm leaving my basket, I'm going to

the competitor. So and I do believe it's a two, it's a

two-systems solution. For most retailers. It's computer vision

that you've been talking about. And I think for things like

apparel, and general merchandise and other stuff, you can run a

camera, buy up a bunch of clothes, but I can't tell one

from another. I think RFID plays a role there as well. So there's

a two, there's a two technology race to figure out how to

leverage this stuff. And I think you guys are well positioned to

do this. Barbara, thank you so much for taking the time out of

your busy schedule. We really do appreciate it. We will be

podcastinh, we will be broadcasting this on both

Conversations on Retail, and the Walton Supply Chain Center in

the next couple of days. So thank you very much for your

time. Any closing comments that you'd like to make before I

before I close off the podcast?

I just wanted to say thank you so much to you,

Mike. I think this is such a an exciting podcast that you have

and it's so relevant and timely with all these technologies that

are available and it's so great that you are taking the lead to

give us another avenue or venue to continue to educate.

Yep, absolutely. Barbara, have a great weekend.

audience. Thank you very much. We'll check you back with you

next time when we're bringing another Conversations on Retail

on on self development availability. Thank you very

much. Bye bye.

Thank you.

Boy, I hope you enjoyed that conversation with

Barbara Chase. Join us next time as we have another industry

professional talking about the capability, process, people and

technology to drive on shelf availability of retail. Have a