Week 43 - Holiday Forecasts Are In: Adobe's $275.1B Online Number, Shopify's New Canvas Tool, and Google Discover Turning Into an AI Concierge

MaxWelcome to the Checkout Point, your quick dive into this week's ecommerce buzz with me, Alex, powered by Blicket. These 20 minutes pack the latest trends, news, and insights. Let's navigate the digital marketplace together. Ready? Let's go.

RachelOkay, confession. I bought an air fryer at 11:58 on the last night of the sale. Did I need it? No. Do I own three already? Also no. Wait, yes.

Speaker 3Try, Max, what's the actual ROI on a fourth air fryer?

RachelZero. Joy, maybe.

Speaker 3Amazon's own page says Prime Big Deal Days ran October sixth to seventh, 48 hours of member deals.

RachelSo Fall Prime Day is done.

Speaker 3Which means the holiday season has officially started, ready or not.

RachelReady? Rachel, I haven't even finished my Halloween candy shopping, and now it's holiday season.

Speaker 3Of course you haven't.

RachelAnyway, I'm Max. This is the Checkout Point, and that's my co-host, Rachel, who will now judge my purchases.

Speaker 3Gladly. The forecasts are rolling in, and honestly, two big ones don't line up the way people think.

RachelPlus, there's an AI question hanging over all of it.

Speaker 3Mm-hmm. Let's get into it.

RachelAll right, here we go. So anyway, the sale is over. Let's talk about what the forecasters say comes next. Okay, numbers. Adobe says US shoppers will spend two hundred seventy-five point one billion dollars online between November first and December thirty-first.

Speaker 3Billion with a B.

RachelBillion with a B. Adobe's release says that's up six point seven percent from last year.

Speaker 3Hmm, and that's a record, I assume, because every forecast is a record?

RachelEvery forecast is a record. Nobody ever releases a press conference titled Pretty Average Year Ahead.

Speaker 3Fair. Give me the sub numbers, though. Where's the money piling up?

RachelAdobe projects Cyber Week at forty-seven point five billion.

Speaker 3Okay.

RachelAnd Cyber Monday, November thirtieth, at fifteen point one billion on its own.

Speaker 3One day, fifteen billion. That's a supply chain bottleneck with a bow on it.

RachelYour warehouse manager just felt a chill.

Speaker 3Mine did, honestly. I've lived through those Mondays.

RachelOkay, story time. What does a Monday like that actually look like from the inside?

Speaker 3Controlled panic. In my retail operations days, the weeks before a peak were all about one question: Can the building physically ship what the website is about to sell? A forecast tells you the demand. It doesn't pick, pack, or drive anything.

RachelSo a big forecast number is really a staffing and inventory question wearing a marketing hat.

Speaker 3Exactly. And a small merchant has a tiny version of the same problem. Fewer boxes, same cliff.

RachelWhat's the cliff?

Speaker 3Selling something you can't ship. Cyber Monday is November thirtieth, which is under two months away. So plan your stock around the products that actually move, not the whole catalog.

RachelAnd plan people. Who's answering the where's my order emails that land the next morning?

Speaker 3Right. The day after is where the pain hides. Sales are fun. Support tickets are the bill.

RachelThe bill always arrives.

Speaker 3Always. Anyway, keep going. What else is in the Adobe coverage?

RachelDigital Commerce 360 covered the Adobe forecast too, and their coverage puts Black Friday at twelve point nine billion.

Speaker 3Wait, so Black Friday is smaller than Cyber Monday in this forecast?

RachelSmaller dollars, yeah, but the coverage says Black Friday is growing faster, up nine point two percent.

Speaker 3Huh. So the Monday's bigger, the Friday's hotter.

RachelRight. Cyber Monday keeps the crown. Black Friday keeps the momentum.

Speaker 3Quick caution on that one, though. That Black Friday figure comes from trade coverage of Adobe's forecast, not from me reading Adobe's own release.

RachelI'd treat it as secondhand.

Speaker 3Good flag. So the headline is one level up, and the Black Friday detail is relayed.

RachelRight. And does faster growth mean better for you? Not automatically. A day that grows faster just means more people crowd into it.

Speaker 3Which could mean more competitors going after the same shoppers.

RachelCould. The forecast doesn't say. I'm only saying don't assume a hotter day is a cheaper day.

Speaker 3Hotter isn't cheaper. Got it. And that goes for Cyber Week too. Forty-seven point five billion is the whole stretch. So a good week for the market could still be a rough week for a single store with thin margins.

RachelA rising tide, but a leaky boat. Okay, I like that, and it's a better metaphor than my gym one.

Speaker 3Okay, now the other forecast, Deloitte.

RachelOh, this is where it gets fun.

Speaker 3Deloitte's September tenth release projects holiday e-commerce at three hundred and sixteen billion to three hundred and nineteen billion.

RachelWait, really? That's forty billion more than Adobe.

Speaker 3Forty-ish, yes, and their growth is higher too, seven and a half to eight point four percent.

RachelSo Deloitte is just more bullish. Somebody's lying.

Speaker 3Nobody's lying. Look at the calendar.

RachelAh, I hate calendars.

Speaker 3Deloitte's release says itself that the window is three months, so it isn't directly comparable to Adobe's.

RachelThat's like comparing my gym attendance in a two-month stretch to someone else's three months and calling it a rivalry.

Speaker 3Your gym attendance is a flat line, Max.

RachelHa, a very stable flat line, low variance.

Speaker 3But seriously, this is how bad decisions get made. Somebody grabs the bigger number for a board deck and sets a revenue goal off it.

RachelAnd the bigger number's got an extra month baked in.

Speaker 3Right. So before anybody quotes either figure, ask what dates it covers.

RachelTotally. Okay, Rachel, you've got a look.

Speaker 3I do, because these are top-line numbers, and top line is not profit.

RachelHmm.

Speaker 3Go on.

RachelSay the whole market grows nearly seven percent. Great. But what's my ad cost doing in the same stretch? What's happening to my margin? Hmm. You mean if everybody's spending, auction prices go up?

Speaker 3I mean, a growing pie doesn't mean your slice is bigger. Market growth tells me nothing about my cost to acquire a customer.

RachelFair. These forecasts can't tell you that.

Speaker 3No. They tell you demand is probably there. They don't tell you what it costs to catch it.

RachelWalk me through it with numbers, Rachel. Pretend I'm a small store, totally made up for illustration only.

Speaker 3Fine. Say you sell a product for a hundred dollars, and after product cost and shipping, you keep forty. That forty is what's left to pay for everything else, including ads.

RachelSo if it costs me thirty to get that customer through ads, I keep ten.

Speaker 3Right. Now say the auction gets more crowded and that same customer costs forty. You made the sale, the dashboard says revenue, and your profit on that order is zero.

RachelRevenue up, profit gone. That's the movie you keep warning about.

Speaker 3And the forecast would still say record year, both things true at once. These are invented numbers, to be clear. Plug in your own.

RachelWhich means the homework is knowing your real per order margin before the season starts.

Speaker 3Before. Then you know the most you can pay to win a customer without losing money on the first order.

RachelSo the number is more like a weather report than a budget.

Speaker 3A weather report with a six-figure ad bill attached.

RachelOuch. And honestly, most small merchants don't have a Deloitte or an Adobe sitting in their dashboard anyway.

Speaker 3Right. They've got their own last year numbers and a nervous feeling.

RachelNervous feeling, which, to be fair, is also a forecasting model.

Speaker 3A low confidence one.

RachelVery low. Okay, so the take is use the forecasts for direction, pick one source and check its dates.

Speaker 3And then look at your own margins before you celebrate anything.

RachelCan we make that concrete? What's a starting point for someone listening?

Speaker 3Write down your own numbers: last year's holiday sales by week, your ad spend by week, and what you kept per order. It's boring. It beats a forecast you can't act on.

RachelAnd then compare the direction of the forecasts to your own history instead of copying their totals.

Speaker 3Yes. Borrow the direction, keep your numbers.

RachelSpeaking of which, there's one number in that Adobe coverage that changes who shows up at your store.

Speaker 3Oh, you're gonna make me ask, aren't you?

RachelLittle bit. It's not a dollar figure at all. Okay, so the number I teased. Digital Commerce 360's coverage of Adobe's forecast says AI traffic to US retail sites is projected to rise a hundred and thirty percent year over year.

Speaker 3Wait, really? Say that again.

RachelOne hundred and thirty. More than double. And it's a projection, so nobody's counted it yet.

Speaker 3Right. Forecasts are forecasts. When does it land?

RachelThe coverage says the biggest increases hit on Thanksgiving, Black Friday, and Cyber Monday. So the peak days.

Speaker 3Of course, exactly when your ad costs are already screaming.

RachelYou can't just let me enjoy a number, can you?

Speaker 3I mean, come on. More visitors sounds great, but what's the actual conversion on a visitor an assistant sent you? Nobody in that coverage says.

RachelFair. The source gives traffic, not conversion. So we know the door's getting busier, not who walks through it.

Speaker 3And for a small store, that distinction is everything. More than double the assistant traffic sounds like a windfall, but you could end up with more sessions in your analytics and the same number of orders. Visits are not revenue.

RachelSo what would you actually watch? If I'm a small store, where do I even see AI traffic?

Speaker 3I can't tell you how any given analytics tool labels it. The coverage doesn't say. What I'd do is check whether your reports separate referral sources at all before the peak days, so you're not guessing in the middle of Cyber Monday.

RachelThat's a good boring task. Check the dashboard before the dashboard matters.

Speaker 3Okay. Hold that thought because there's a second piece.

RachelSearch Engine Journal, October twelfth. Roger Montti reported that Google is planning to turn Discover into an AI concierge.

Speaker 3Concierge, like a guy in a vest who recommends the restaurant?

RachelKind of, yeah. The stated goal is making Discover more useful.

Speaker 3Useful to whom?

RachelThat's the catch. The same report flags the risk, which is less traffic for publishers.

Speaker 3Hmm. So the helpful thing and the traffic-killing thing are the same thing?

RachelPretty much. Though the report is about publishers, I won't pretend it says anything about your product pages.

Speaker 3Right, but merchants live next door. A lot of us lean on content to pull people in.

RachelTotally. So here's where my brain went. AI sends you a flood of visitors. Great. Reach goes up.

Speaker 3And the same kind of system decides what gets shown at all.

RachelYes. Whoever sits between you and the shopper picks the shelf.

Speaker 3Which is my supply chain brain screaming. You've got a distributor who can reroute you whenever they want.

RachelOkay, but is that worse than before? Google and Amazon already sorted the shelf.

Speaker 3Sure, but a ranked list of ten links is different from one answer. If the assistant names one or two stores, the rest of us are invisible.

RachelHuh. That's a fair difference. Ten shots versus two.

Speaker 3And to be fair to the other side, nobody has shown that's how it plays out. The Discover report describes a plan, not a finished product. Plans change, and the stated goal is to be more useful to the people using it.

RachelSo the honest version is it's a direction, and you get to decide how much to hedge against it.

Speaker 3Right. Hedge is the word. I'm not saying panic and rip out your content. I'm saying don't let one door be the only way anyone finds you.

RachelWhich is a very nice way of saying diversify without using the word diversify.

Speaker 3So what does that do to unit economics? If you're one of the two, maybe your cost per sale looks great.

RachelAnd if you're not?

Speaker 3Then you pay more elsewhere to make up for it, or you pay the platform to be seen. Either way, the margin shrinks.

RachelI can't prove that from these sources, to be clear. It's the risk, not a reported result.

Speaker 3Agreed. It's the question to ask before the forecast makes you hire three people.

RachelWalk me through the unit economics version with the caveat that it's an illustration and not a finding.

Speaker 3Sure, and it's illustration only, not from any source. Take your margin per order, then ask what happens if you have to spend more to get that same order, either because a channel got more crowded or because you're paying to be shown. If your margin can't absorb that, the traffic isn't growth, it's a bill. And the real test is how many of your orders come from places you can't turn off or negotiate with.

RachelThe forecast hired three people. Classic.

Speaker 3Because if the gatekeepers keep shifting, who actually owns the customer?

RachelYeah, because right now the answer might not be you. Okay, so who owns the customer? I found a piece that tackles exactly that. Mexico Business News ran it on October 12th, titled Who Actually Owns Your Customer in Mexican E-Commerce. Oh, I like that title. Straight to the point.

Speaker 3Right? The writer is Juan Martn Vignart, and he argues that direct-to-consumer E-Commerce in Mexico builds long-term customer relationships, retention, and value beyond a marketplace transaction.

RachelBeyond the transaction, meaning a marketplace sale is a one-and-done?

Speaker 3Careful. The summary I have says what he argues, not a data set. It's a case for D2C, and I'd treat it as an argument, not proof.

RachelFair. Still, it's a clean idea.

Speaker 3It is, and it's my lens on everything. A marketplace sale gets you revenue today, but what's the actual lifetime value of a shopper who bought from you through someone else's storefront?

RachelYou didn't even wait a beat on that one.

Speaker 3Because if the marketplace holds the email, the reviews, the whole relationship, you can't bring that shopper back. You're renting them. Retention is a supply chain problem, Max. If you don't control the inventory of customers, you can't restock.

RachelAh, okay. I'll give you that one.

Speaker 3Thank you.

RachelSo how do you actually own the relationship? Because build a brand is a nice slogan. What's the plumbing?

Speaker 3First party data. Who's on your list, what they bought, what they do on your site.

RachelAnd that's where tracking comes in, which is unglamorous but matters. A GlobeNewswire release says Elevar published a Captain Compliance setup for Google Consent Mode and server-side tracking on Shopify.

Speaker 3Okay, translate for the people who just heard four nouns in a row.

RachelFair. Consent Mode is how your tracking respects what shoppers agreed to. Server-side tracking means the data flows through your own setup instead of only the browser. And Captain Compliance is the consent layer Elevar says it documented for Shopify stores.

Speaker 3And it's a press release, so that's the company's own description of its own work.

RachelTotally. I'm not telling anyone it fixes their attribution. I'm saying it shows where the plumbing conversation is going.

Speaker 3Which is the right conversation. If shoppers arrive through AI surfaces you don't control, your own data is what's left.

RachelWait, really? That's the whole connection? Consent plus tracking equals owning the customer?

Speaker 3No, it's a piece of it. Consent gets you permission. Tracking gets you signal. Retention is what you do with them afterward.

RachelRight. And honestly, as a CRO nerd, I've seen a lot of first-party data plans. Most of them are a spreadsheet named Final V Seven that nobody has opened since March.

Speaker 3I have personally been in that spreadsheet.

RachelEverybody has. It's the haunted house of marketing.

Speaker 3But seriously, tie it back. If you're a merchant, ask one question. When this shopper buys again, do I get to be the one who remembers them?

RachelOoh, that's a good test. Can you email them? Can you recognize them? Can you tell if they came back?

Speaker 3And can you do it legally, which is the consent part. Skip that, and your clean data becomes a liability.

RachelMm, yeah.

Speaker 3Cheap reminder, though. One source here is an argument, and the other is a company's own description. I'm using them as prompts, not proof.

RachelGood reminder. They're conversation starters, and the real work is on your side of the screen.

Speaker 3So no, it's not glamorous, but it's cheaper than buying the same customer back through ads every single time.

RachelOkay. Give people a checklist, something they can start without a developer.

Speaker 3Fine. One, know where your customer list actually lives. Is it in your store, in an email tool, or in a marketplace account you can't export from?

RachelTwo, make sure people can clearly say yes to hearing from you right at checkout.

Speaker 3Three, check that your orders connect to a person so you can tell a repeat buyer from a first-timer.

RachelFour, find out what your tracking does when a shopper says no to being tracked. That's the consent piece, and it's worth a conversation with whoever set it up.

Speaker 3Five, pick one retention number and watch it on a schedule. Repeat purchase rate or how many people come back. Just one.

RachelNot 12 dashboards.

Speaker 312 dashboards means nobody looks at any of them.

RachelPainfully true.

Speaker 3None of this needs a vendor to start. It needs someone to own the list.

RachelSpeaking of cheaper ways to get things done, before I forget, there's new tooling news that might help with all of this.

Speaker 3Oh, the Shopify thing?

RachelThat's the one. Okay, tooling news. Practical Ecommerce put out its weekly merchant tools roundup dated October seventh, and three items jumped out.

Speaker 3Only three? That's restraint.

RachelI'm learning. First one, Shopify launched Canvas. It's a design surface where you build layouts with Sidekick, which is Shopify's AI agent.

Speaker 3So you talk to the robot, and the robot lays out your homepage.

RachelRoughly, yeah. The roundup says merchants build layouts with Sidekick on Canvas. That's all I'm gonna claim about how it works.

Speaker 3Mm, because I have no idea if it's fast or good.

RachelRight. No results in the roundup. Nothing on conversion, nothing on time saved.

Speaker 3Which is my whole problem with launch news. A new design tool is a new thing to learn, and learning has a cost.

RachelWait, you're gonna tell me this is a supply chain thing?

Speaker 3Everything is a supply chain thing. If your team spends the two weeks before the rush redoing layouts, you've moved the bottleneck, not removed it.

RachelOkay, fair. Though if a layout is stale, you probably wanna fix it before traffic shows up.

Speaker 3Sure. Before, not during.

RachelAnd if your layout is fine, nobody's forcing a redesign just because something shiny launched.

Speaker 3Right. A launch date on their calendar isn't a deadline on yours, especially in the middle of peak season.

RachelTotally. Second item: Swap acquired Vizby, and Vizby is an AI-visibility tool.

Speaker 3Hold on. Visibility into what?

RachelThe name says AI visibility and the roundup stops there. I'm not gonna make up what the dashboard shows.

Speaker 3Fine, but it's interesting that it exists. Earlier, we were talking about AI as a gatekeeper. Somebody's selling a way to see how you show up there.

RachelSomebody bought one even.

Speaker 3Right. Acquisitions are a signal that people think the problem is real. They're not proof that the tool works.

RachelMm-hmm. Big difference.

Speaker 3Huge. I'd wanna know what it costs and what I'd do differently on a Tuesday because of it.

RachelThat's your question for everything.

Speaker 3But what's the actual ROI? Yes, it is.

RachelI knew it was coming. Third one: MoEngage launched a native Shopify integration.

Speaker 3Native as in built to plug in, not duct taped together.

RachelThat's the idea behind the word, but the roundup just says native integration. Whether it's smooth for your store, I can't tell you.

Speaker 3I'll say this: anything that connects messaging to your Shopify data without a pile of custom work is at least pointed at the right spot because that's where repeat purchases come from.

RachelOkay, that tracks with the retention stuff we just did.

Speaker 3It does. Still a pointer, not a promise.

RachelSo three launches, zero performance numbers. If you had to rank them for a store that's about to get slammed.

Speaker 3Hmm. Honestly, none of them before the rush, unless one fixes something already broken for you.

RachelOoh, even Canvas?

Speaker 3If your layout is fine, leave it alone. If your email tool is the thing that's limping, the MoEngage piece might be worth a look. If you've got no clue how AI sees your catalog, Vizby's category is at least worth knowing about.

RachelLet me push on that. How would somebody test a tool without wrecking their week?

Speaker 3Four questions: What specific problem is it for? What will it cost in money and in your team's hours? How will you know it worked? And how do you undo it if it doesn't?

RachelThe undo one is underrated.

Speaker 3Because holiday traffic is a bad time to find out something can't be rolled back. If you can't picture reversing it, wait until January.

RachelAnd how will you know it worked needs a metric before you start, not after.

Speaker 3Exactly. Decide the metric first. Otherwise, you look at whatever went up and call it a win.

RachelEverybody's done that.

Speaker 3Turns guilty. And one more: Who owns it? A new tool without an owner is a subscription nobody opens.

RachelSo you try it small, one page or one email flow before trusting it with everything?

Speaker 3If at all. The roundup gave us no results, so the burden of proof sits on you.

RachelSo the filter is what's broken right now.

Speaker 3Yep. Shiny is not a diagnosis.

RachelPut that on a mug. All right, let's wrap this up quickly with one thing to do this week.

Speaker 3Just one. I like it.

RachelOkay, so the wrap. Quick scoreboard before we let everyone go.

Speaker 3Quick being the keyword. Two forecasts that can't be lined up side by side, a projection of AI sending more shoppers, and a lot of talk about who owns the customer.

RachelThat's the whole episode in one breath. Impressive.

Speaker 3I've had practice. Budget meetings at Amazon will do that to you.

RachelSo what's the homework, Rachel?

Speaker 3One action. Pull up your holiday plan and mark every piece that depends on a channel you don't control: ads, marketplaces, search, and AI assistant you can't see inside.

RachelMark it how?

Speaker 3Highlighter. Literal or digital, I don't care. Then look at how much of your plan is highlighted.

RachelOh, that's gonna be a lot of yellow.

Speaker 3For most people, yes. And then ask the ROI question. If one of those channels changes its rules in November, what happens to your margin?

RachelNobody wants to answer that in November.

Speaker 3Which is why you answer it now while it's still a spreadsheet and not a panic.

RachelGive me a concrete first step. What does someone do Monday morning?

Speaker 3Open the plan, list the channels, and write one sentence next to each. If this changed tomorrow, what would I do? One sentence. If you can't write it, that's your risk, and now you know where it is.

RachelTotally. And hey, if your plan comes out looking like a yellow school bus, that's fine. Just know who's driving it.

Speaker 3Max, that's almost wise.

RachelI have my moments, maybe two a year.

Speaker 3If you want more of this, head to blicket.co and subscribe so the next episode just shows up.

RachelDo it. Future you will thank you. Thanks for hanging out with us at the Checkout Point.

Speaker 3Go check your highlighter supply.

RachelAnd be kind to yourselves out there. Happy holiday prep, everyone.