by Michaela Sharp | Oct 12, 2026 | Dalilu Wire, Digital, Programmatic
Amazon has introduced DVA+, a new AI-assisted programmatic buying experience. The kid is new, but its parent is anything but.
There’s a new kid on the programmatic advertising block, and this one comes from a rather well-established family.
Its name is DVA+, and its parent is Amazon.
Now, Amazon is hardly a newcomer to digital advertising. It already operates a substantial advertising technology business, sells advertising across its shopping and entertainment properties, and gives advertisers access to inventory well beyond its own platforms.
On September 29, 2026, Amazon announced DVA+, a new AI-assisted buying experience within Amazon Ads Agent, its unified advertising platform.
DVA+ isn’t an entirely new demand-side platform (DSP). It’s a different way of creating and managing display, video and audio campaigns using Amazon’s existing programmatic capabilities.
And what caught my attention wasn’t simply another advertising acronym. It was how familiar the whole process sounded.
A Little Less Complicated, Please
If you’ve ever built an advertising campaign in Meta, you might recognize the approach.
Choose your objective, upload your creative assets, set your budget, and let the platform do much of the work. The system takes care of targeting, bidding and optimization, with varying degrees of control available to the advertiser.
DVA+ introduces a similar concept to programmatic advertising. According to Amazon, its AI can help manage audience selection, bidding, budget allocation and campaign optimization across multiple advertising formats, including streaming TV, online video and display, with audio capabilities expanding as the product rolls out.
I can certainly appreciate the appeal. Not every marketer wants to spend their day navigating complicated DSP interfaces, adjusting settings and studying delivery reports. Some prefer to give the platform an objective and let the technology figure out how to get there.
Personally, I prefer a hands-on approach.
I enjoy selecting inventory, working with audiences, activating private marketplace deals and making adjustments as a campaign progresses. There’s something satisfying about watching the data come in, recognizing an opportunity and knowing which lever to pull next.
That’s why one aspect of Amazon’s announcement particularly interests me.
DVA+ isn’t designed exclusively for advertisers who want to hand everything over to automation. Amazon says its advanced settings also allow more control over targeting, bidding, frequency management, inventory selection and private marketplace deals.
For me, the ideal combination is technology that simplifies repetitive work while giving me the freedom to make informed decisions. AI can analyze data, identify patterns and suggest adjustments, and I’m happy to have that assistance.
But I still want the final say in how my campaigns are managed.
Now, Let’s Meet the Family
Here’s where this new arrival gets particularly interesting.
DVA+ doesn’t come from a small advertising technology startup trying to establish itself in a crowded marketplace.
Its parent is Amazon.
And Amazon has built quite an advertising family.
There’s the shopping marketplace, where consumers search for products, compare options and make purchases.
There’s Prime Video, offering access to streaming entertainment and advertising-supported viewing.
There’s Twitch, with its live-streaming communities, and Fire TV, connecting Amazon with audiences through their television screens.
Then there’s Amazon’s programmatic advertising infrastructure, which provides access to inventory across Amazon properties and third-party websites, apps and streaming services.
Not every advertising environment accessible through Amazon belongs to Amazon, of course. Many are independent publishers and advertising partners.
Still, when you start looking at everything connected to Amazon’s advertising business, you begin to appreciate just how extensive that ecosystem has become.
And that brings us to something that interests me even more than the campaign setup.
The Family Knows Quite a Bit About Its Customers
Consider what happens when someone visits Amazon to browse for a new coffee machine.
They might compare models, read reviews, check prices and eventually purchase one.
Amazon can observe those interactions within its own shopping environment.
Now consider the shopping, browsing and entertainment signals available across its services, together with the advertising opportunities accessible through its technology.
That’s a substantial collection of information to work with.
For advertisers, the potential to connect shopping interests with advertising across multiple formats is particularly interesting.
Of course, consumer behaviour is more complicated than a shopping history. Someone researching coffee machines isn’t necessarily ready to purchase one, and not every advertising objective revolves around selling products online.
But Amazon’s combination of shopping, browsing and streaming signals provides an interesting foundation for its advertising technology.
How effectively DVA+ puts those capabilities to work is something we’ll learn as more advertisers begin using it.
So, When Can We Meet the New Kid?
There’s one important detail Canadian advertisers should know.
DVA+ was introduced in an extended closed beta in the United States, with an open US beta expected in late October 2026. Amazon has announced plans for international expansion during the final quarter of 2026, followed by general availability across 35 markets targeted for the first quarter of 2027.
That doesn’t mean Canadian advertisers are waiting for Amazon’s advertising inventory to arrive. Amazon’s existing DSP capabilities and Canadian inventory are already available.
What’s new is the DVA+ buying experience, and its Canadian launch timing has not yet been specifically confirmed.
So, for now, we’re looking at an interesting development rather than a platform most Canadian media buyers can immediately sit down and test.
So, What Do I Think of the New Kid?
What interests me most about DVA+ isn’t simply the promise of making programmatic advertising easier. It’s the combination of automation, professional buying controls and the advertising ecosystem behind it.
But I should make something clear. Everything I’ve shared here is based on Amazon’s announcements and what I’ve read about the new experience. I haven’t tested DVA+ myself, so these are my initial observations, not a hands-on review.
And as a media planner, that’s where things get interesting for me.
Once I have access to it, I might just roll up my sleeves and put a campaign or two through the platform. I’d like to see how it handles inventory selection, optimization, reporting and, most importantly, how much flexibility it actually gives me.
If I do, I’ll share what I discover, including what works well and what might need a little more attention.
Until then, I’d love to hear what other media professionals think. Does this kind of automation appeal to you, or do you prefer keeping more control over your campaigns? And if you’ve already had an opportunity to explore DVA+, what has your experience been?
After all, there’s a difference between reading about what a platform promises and actually sitting down to use it.
The new kid has caught my attention. Now I’d like to see how it performs.
Dalilu Wire | Media Decoded
Industry commentary based on Amazon Ads announcements from September 29, 2026. DVA+ was in a limited US beta at the time of writing. Product capabilities and geographic availability may change.
Primary source: https://advertising.amazon.com/en-ca/resources/whats-new/dva-plus-display-video-audio-campaigns
by Michaela Sharp | Oct 10, 2026 | Dalilu Wire, Digital, Media Buying, Programmatic
Five Canadian broadcasters are exploring a shared connected-TV marketplace. There’s plenty to like about the idea. But will media buyers gain more flexibility, or will advertisers be asked to compromise?
MEDIA & ADVERTISING ANALYSIS | October 2026
I’m all for making Canadian media stronger. And if five of our largest broadcasters can work together to create better advertising opportunities, that’s something worth paying attention to.
But I’m also a media buyer. My responsibility is to build campaigns that work for my clients, make the most of their budgets and give me the flexibility to adjust when needed.
So when I hear about a proposed new connected-TV (CTV) marketplace, my first question isn’t how many broadcasters are involved.
It’s whether I’ll be able to buy that inventory through the DSPs I already use.
Because the last thing I want is another standalone platform that makes me separate budgets, reporting and campaign decisions that I’d much rather manage together.
I’m crossing my fingers this new initiative will be an advantage for media buyers and their clients, not something that requires either of us to compromise.
And that brings us to an interesting announcement.
Five broadcasters, one potential marketplace
CBC/Radio-Canada, Bell Media, Corus Entertainment, Rogers Sports & Media and Quebecor are exploring a joint initiative tentatively called the Canadian Coalition for Connected TV (CCCTV).
The proposal was discussed at the Future of TV Advertising Canada conference in Toronto on September 29, 2026, and reported by Canadian Ad Insider on October 1.
The concept is to create a shared gateway through which advertisers could access premium CTV inventory from all five broadcasters.
For the broadcasters, there’s an understandable business case. They’re competing for advertising dollars in a market dominated by powerful international technology companies, while also investing in Canadian programming and audiences.
Working together could help strengthen their offering and keep more advertising investment in Canada.
That sounds promising.
But whether it delivers equal value to the people actually buying the advertising is another question.
Buying media isn’t the problem. Buying it efficiently matters.
Let’s consider how a typical digital campaign works.
We might be running display advertising, online video, CTV and several audience-targeting strategies, all with different objectives and performance expectations.
Through a demand-side platform (DSP), we can manage multiple tactics, evaluate inventory, adjust budgets and optimize campaigns as results come in.
We also already have ways to access premium inventory through Private Marketplaces (PMPs) and Programmatic Guaranteed deals.
Finding inventory, evaluating opportunities and negotiating media deals are part of our job. We don’t necessarily need someone to eliminate that work.
What we need is the freedom to make the best media decisions for our clients.
And that includes being able to move budgets between tactics when the campaign tells us something needs to change.
Now imagine being required to purchase part of that campaign through a separate platform.
Separate reporting dashboards. Budgets managed in different places. Additional steps to bring campaign results together.
The measurements themselves may be comparable, but managing and reviewing the entire campaign could become less convenient.
That may be manageable, but is it an improvement?
The DSP question
This is where CCCTV could become a genuinely useful addition to Canadian programmatic advertising.
If the coalition makes its inventory available through established DSPs, media buyers could potentially incorporate those opportunities into their existing campaign workflows.
That would be welcome.
But if buying requires a separate platform, advertisers and agencies may face trade-offs in budget flexibility, reporting and coordinated optimization.
DSP access alone doesn’t guarantee unified frequency management or reporting. Those capabilities depend on the technology, data and integrations behind the marketplace. Still, DSP compatibility would be a meaningful starting point.
And while we’re asking questions, there are a few others worth considering.
Will the coalition deliver incremental Canadian audience reach? Will it offer competitive pricing and meaningful inventory packages? Will smaller independent agencies have access to the same opportunities as larger media-buying organizations?
These aren’t objections to the initiative. They’re the practical questions that determine whether something works in the real world.
We don’t have those answers yet
According to Canadian Ad Insider’s October 1 reporting, the coalition remains in the early stages of development.
No official technology partners had been announced at the time of that report, although preliminary discussions with supply-side platforms (SSPs) had reportedly taken place.
That means we shouldn’t assume CCCTV will become another walled garden. It may very well develop into a marketplace that works within existing programmatic buying environments.
And I hope it does. Because there’s a real opportunity here.
Let’s make this work for everyone
Canadian broadcasters have every reason to explore ways of strengthening their position in the advertising market.
Better access to Canadian programming and audiences could benefit advertisers, agencies, broadcasters and the Canadian media industry more broadly.
But collaboration alone won’t determine whether this initiative succeeds.
The real test will come when media buyers can evaluate the inventory, integrate it into their campaigns and determine whether it delivers value for their clients.
I would love to see CCCTV become a strong new option within the DSPs we already use. Give us competitive inventory, meaningful audience opportunities and the flexibility to manage it alongside our other digital investments.
That’s the kind of progress I can get behind.
Because creating another place to buy advertising isn’t necessarily the same as making advertising easier to buy.
And if the goal is to attract more investment into Canadian media, let’s make sure we’re giving advertisers more reasons to invest, not more compromises to make.
Background: Canadian Ad Insider, October 1, 2026; Future of TV Advertising Canada, September 29, 2026.
by Michaela Sharp | Oct 7, 2026 | Blog, Dalilu Wire, Digital, Programmatic
I had a meeting recently with a media representative from a TV and radio network.
Like many reps in our smaller Canadian market, digital is only one part of the portfolio. The same representative may also be selling TV, radio, sponsorships and other media products.
During the conversation, I mentioned a PMP.
The term wasn’t familiar.
That stayed with me.
Not because I expected this person to be a programmatic specialist. In a smaller market, media representatives often need to know a little bit about a lot of things.
But it did remind me how fragmented the understanding of programmatic advertising still is. If you worked only in digital or programmatic, PMP would likely be familiar territory. For someone selling across several media channels, maybe not.
So let’s decode one small piece of it.
First: What Is a PMP?
PMP stands for Private Marketplace.
It allows publishers and inventory providers to make selected digital inventory available programmatically to specific buyers, usually through a Deal ID.
| Publisher / Inventory Provider → SSP / Exchange → PMP / Deal ID → DSP → Advertiser |
Kaboom! Did your brain just explode? No worries. If you stick with me, I’ll slowly introduce you, article by article, to all of these scary-looking acronyms.
For now, all you really need to know is that a PMP gives me a way to access a particular package, publisher, placement or product that I specifically want.
Reason #1: I Want Access
Here in the Canadian market, I might be interested in inventory from CBC, Corus or Bell Media. Or I might be talking to an exchange or another inventory provider about something much more specific.
Right now, for example, I am talking with one provider about Connected TV opportunities around particular sporting events and a ticker product.
What do you have? Where could it fit into my media plan? And can I buy it programmatically?
If the product makes sense, the vendor can make it available to us through a PMP.
So Where Does the Deal Go?
This is the slightly technical part.
Dalilu has a buyer seat in the programmatic ecosystem. That seat has an identifying number.
We give the appropriate seat information to the vendor, which tells them where to send the deal. They make the PMP available to us, we receive the Deal ID, and the inventory becomes accessible through our DSP.
From there, I can activate it. And more importantly: I can manage it myself.
Reason #2: I Like Control
At this point you might be thinking: she is a total control freak.
YES. When it comes to my clients’ campaigns, I absolutely am.
If I am interested in a vendor’s digital product, that doesn’t necessarily mean I want to hand them part of my client’s budget and have their team manage that portion of the campaign separately.
I want access to their product. Then I want to decide how it fits with everything else I am buying.
I Don’t See Four Campaigns. I See One.
A client campaign might include several programmatic tactics. A PMP might be one. Video might be another. Display or mobile might be another.
But I don’t look at those as completely separate campaigns. To me, they are one campaign.
I want to see the budget as a whole. I want to see how the tactics are working together. I want to optimize them together. And I want to manage frequency across them as intelligently as possible.
The Consumer Doesn’t Care Who Bought the Ad
This is where things can get messy when several vendors are managing separate pieces of a digital campaign.
Vendor A might reach someone four times. Vendor B might reach them another four times. Vendor C might add another four.
Individually, each campaign may look perfectly reasonable. Together? That person has now seen the client’s advertising twelve times.
The consumer doesn’t know which vendor served which impression. They just know: I have seen this ad a LOT.
By managing those tactics together in the DSP, with cross-device targeting helping us connect exposures across devices, I can manage frequency much more holistically. That gives me better control over the campaign as a whole and helps reduce unnecessary overexposure.
PMPs Give Me Access Without Giving Up Control
And that, for me, is really the appeal.
A PMP can give me access to inventory I specifically want without forcing me to give up control of the larger media strategy.
I can negotiate the inventory and the deal directly with the vendor, bring it into the DSP, combine it with other tactics and inventory sources, and then manage and optimize the campaign as one connected plan.
It also means I am not relying too heavily on one media owner or one channel. I can build broader reach by bringing different pieces together.
One Last Thing: “Private” Does Not Mean “Perfect”
A PMP does not automatically mean better inventory.
Put mediocre inventory behind a Deal ID and it does not magically become premium.
The same questions still matter: Who are we trying to reach? Where is the advertising actually running? What are we paying? How much scale is available? What frequency are we building? And most importantly: does this inventory actually make sense for the campaign?
You don’t need to become an ad-tech engineer to understand programmatic advertising. But if you are responsible for spending marketing dollars, understanding a little bit of the plumbing can be very useful.
by Michaela Sharp | Oct 7, 2026 | Dalilu Wire, Digital, Media Buying
On a media plan, “programmatic” can look remarkably simple: one line, one budget, maybe a few tactics underneath it. Behind that line, however, a lot is happening. Bids are changing, inventory is being evaluated, domains are being watched, ads are being compared, budgets are moving, patterns are emerging and problems are being fixed.
All of that work happens inside something called a DSP.
DSP stands for Demand-Side Platform. Very simply, it is the technology media buyers use to buy, manage and optimize programmatic advertising. I think of it as the control room of a digital campaign. Having access to that control room is one thing. Knowing what to do once you are inside it is another.
There Is More Than One DSP
There are many DSP providers available to agencies and media buyers, and each offers its own mix of inventory connections, targeting capabilities, data, reporting, optimization tools and technology.
The platform matters, but the platform does not make the media buyer. Two people can have access to the same DSP and run very different campaigns because the technology gives you the controls; it does not tell you which ones to use.
Launching Is the Easy Part
Before the first impression is ever served, there are already a lot of decisions to make. Who are we trying to reach? Where? On what devices? Through which inventory? How much are we willing to bid? How often do we want someone to see the message? Which creative should run where? And how should all of those pieces work together?
Then the campaign launches.
That is not the end of the work. That is when the campaign starts talking back.
When the campaign starts talking back, I start listening.
What Do I Look at First?
One of the things I like about working inside a DSP is that there is no single metric that tells me whether everything is working.
It depends on the tactic.
With display advertising, CTR is one of the first clues I look at. If it is below benchmark, I start digging. I look at page loads, individual domains, exchanges and creative to understand what is actually happening.
And sometimes CTR looks fairly ordinary while page loads suddenly spike.
That gets my attention.
Because if people are reaching the website in meaningful numbers, the campaign may be doing considerably more than the CTR alone suggests. To me, that can be gold.
Video is different.
With short outstream video, I look at both CTR and video completion rate. I want to understand whether people are watching and whether the format is also generating response.
With instream video and Connected TV, completion rate becomes my number-one signal. Those formats are primarily there to be watched, so I want to know whether people are actually getting through the video.
The point is not to find one favourite metric and apply it to everything.
It is to know which signals matter for the tactic you are buying.
Then I Go Down a Layer. And Another.
I look at individual domains, especially the ones delivering a large share of the impressions. If one domain is delivering heavily but producing very little response, I click down a layer and look at which exchange is supplying those impressions. Then I click down again and look at which ad is being served.
That is where it can get interesting.
I might discover that one big-box ad coming through one exchange is barely moving, while the same ad size coming through another exchange is performing beautifully. That is my gold ticket.
Now I have something actionable. I might raise the CPM bid because I want more access to what is working. And yes, it works the other way too. If you are not performing, my dear, you may be out and paused.
Or perhaps I do not kick you out completely. I lower the bid and move you a little farther back in line.
Yes, I talk to my campaigns.
What Optimization Really Means
Optimization is one of those digital-media words that sounds much more mysterious than it needs to. Sometimes it means digging through domains, exchanges, creative, bids, delivery and response until a useful pattern becomes clear.
Then comes the judgment call. Give this one more room. Pull that one back. Pause that. Raise this bid. Lower that one. Or do absolutely nothing.
The DSP gives me the data and the controls. It does not make those decisions for me.
Sometimes Part Pilot, Part Detective, Part Media Mechanic
The control-room comparison is actually pretty accurate.
Some days I am the pilot, watching the campaign from a broader level. Is the budget pacing properly? Are the tactics working together? Are we delivering where we should be?
Other days I am the detective, three layers down looking at a domain, an exchange and a specific creative because something does not quite add up.
And every now and then I become a slightly obsessive media mechanic, raising a bid here, lowering one there, giving one tactic more room and pulling another one back.
That is the part I love.
Finding Efficiency Is the Fun Part
What I enjoy most about DSP work is finding efficiencies, seeing patterns and fixing problems. Sometimes a pattern jumps out immediately. Sometimes it takes time.
Clients rarely see most of this work, and I do not think they should have to. They should be able to understand the media strategy, know what we are trying to accomplish and receive reporting that makes sense.
Behind that simplicity can be a great deal of watching, digging, comparing and adjusting.
The complexity stays on our side of the table. We make it easier on theirs.
One Campaign Can Have a Lot Happening Inside It
A programmatic campaign may include Connected TV, online video, display, mobile and perhaps inventory accessed through a PMP.
If PMP is another acronym currently making your eyes glaze over, I wrote an article about that too.
Those different tactics may be doing different jobs, but I still want to understand how they work together. What is building reach? What is reinforcing the message? Where is the budget going? What is being overused? What needs more room?
That broader view is one of the reasons I like working inside a DSP. It lets me see the pieces without losing sight of the campaign.
The DSP Is Powerful. It Is Not the Strategy.
A DSP can buy Connected TV, but it cannot tell me whether Connected TV belongs in the media plan. It can access audiences, but it cannot tell me which audience matters most to the business. It can accept a PMP deal, but it cannot decide whether that deal is strategically worth buying.
Technology can execute a media strategy brilliantly. It cannot replace one.
Strategy first. Everything else follows.
For the broader programmatic picture, see Programmatic Advertising: Knowing It Exists Is Only the Beginning.
One Line on a Media Plan. A Lot Behind It.
So the next time you see “programmatic” sitting neatly on a media plan, remember that the line itself is the simple part.
Behind it, someone should be watching, questioning, comparing, finding patterns, moving budgets, adjusting bids, fixing problems and occasionally deciding that the best thing to do is nothing at all.
You do not need to know how to operate a DSP yourself. But if programmatic advertising is part of your media plan, it is worth knowing who is sitting in the control room.
by Michaela Sharp | Sep 26, 2026 | Blog, Dalilu Wire, Digital, Media Planning
AI has changed the way we search, ask questions and make decisions. Now advertising is following us there and that creates a very interesting new place for media planners to work.
A note on timing: This article was written in September 2026. AI is evolving extraordinarily quickly, and some of the platforms, advertising options, usage numbers and capabilities mentioned here may look very different six or twelve months from now. Consider this a snapshot of where we are today.
If you are responsible for a media plan right now, you are probably hearing about AI from every direction. Your team is using it, your agency is talking about it, and someone has probably already asked whether your brand should be doing something with it.
Until recently, that conversation mostly belonged on the creative, productivity or technology side. Now it has entered media planning, because ChatGPT has advertising and that makes the conversation more interesting.
Not because every marketing director suddenly needs a ChatGPT campaign, but because another environment has opened up where people are asking questions, comparing options and trying to make decisions. That is usually where I start paying attention.
For us, this is no longer something to simply watch. We are already looking at where ChatGPT advertising can fit into media plans, especially as a lower-funnel tactic when the audience, message and level of investment make sense.
The strategic takeaway is simple: do not add ChatGPT because it is new. Add it when it gives the campaign access to a useful moment of intent that the current media mix may be missing.
There was a moment when ChatGPT felt like a novelty. People tried it, talked about it and wondered whether it would really become part of everyday life.
That moment is over. For many of us, ChatGPT is now simply another icon sitting on the phone or tablet right beside the other apps that quietly became part of our routine.
We ask it to rewrite an email, plan a trip, explain something we do not understand, compare products, help with a recipe, work through a purchase or brainstorm an idea. Sometimes we even use it to settle an argument at the kitchen table.
In other words, it has moved in. And whenever human behaviour changes, media eventually changes with it. Advertising has now followed us into ChatGPT.
This Is No Longer a Small Audience
The numbers tell the story rather quickly.
What I find particularly interesting is that this is not only a young-person story. Older adults are becoming part of the AI audience too.
You can easily imagine a grandmother asking ChatGPT to explain a term in simpler language, help her write a birthday message, compare winter boots, suggest what to cook with what she already has in the fridge, or explain how to change a setting on her phone.
For some people, this may simply feel friendlier than traditional search. Instead of typing a few words into Google, opening several websites, closing pop-ups and trying to figure out which answer actually applies, you can ask a question and keep the conversation going from there.
It does not replace human connection, of course, but it is changing how people interact with information.
Yes, I Hear the Concerns
AI will make us lazy thinkers. It is dangerous. We will rely on it too much. Yada, yada, yada.
Some of those concerns are absolutely fair. But from a marketing perspective, debating whether people should use AI is becoming less useful than understanding how they are using it.
The more interesting questions are what habits will stick, what people will start asking AI that they once asked Google, where product comparison will happen, and where decisions will begin. Those are media questions.
We Used to Search. Now We Ask.
For years, the pattern was fairly simple. If people wanted information, they searched. If they wanted inspiration, they scrolled. If they wanted entertainment, they watched. Now we are increasingly doing something slightly different: we are asking.
From a media-planning perspective, that difference matters. A search engine gives you results. A social feed gives you content. A conversation with AI gives you context.
What kind of flooring would work best in my basement?
Which accounting software makes sense for a small business?
Where should we go for a family vacation?
What should I look for when buying a new mattress?
Which vehicle would work for a family of five?
That person has already told the platform quite a lot about what they are trying to accomplish. They have intent, and intent is very interesting to advertisers.
ChatGPT Is Not the Only AI in the Room
Of course, ChatGPT is not the only AI platform people are using.
Google says the Gemini app passed 1 billion monthly users in August 2026. Claude is much smaller, but growing quickly, particularly among work and technical users. Similarweb also tracks Perplexity, Grok and Microsoft Copilot as part of an increasingly fragmented AI landscape.
In other words, this is not a one-platform story anymore.
People are beginning to develop preferences. Some use ChatGPT. Some use Gemini. Some use Claude for work. Others move between several tools depending on what they are trying to do.
And that matters for marketers because the AI audience is starting to spread across different environments, just as it did with search, social and streaming.
So Why ChatGPT?
Because this is where the media-planning opportunity becomes very tangible.
OpenAI now offers direct advertising inside ChatGPT, with ads appearing separately from the AI response for eligible users on the Free and Go plans. Canada is currently available for self-service advertising through OpenAI Ads Manager.
That gives us something very practical to work with: not simply an AI platform to understand, but an actual media placement we can evaluate, plan, buy and measure.
And Now There Is an Ad
Advertising in ChatGPT is now available in Canada through OpenAI’s Ads Manager.
There is, however, an important distinction here: brands are not buying the answer.
Advertising does not change what ChatGPT tells you. Ads remain separate from ChatGPT’s responses and are clearly labelled as sponsored placements. A user asks a question, ChatGPT provides an answer, and when there is a relevant advertising opportunity, a sponsored placement can appear separately below the response.
This is the part I find particularly interesting as a media buyer. The message can appear while someone is already thinking about a subject, exploring options or moving toward a decision. That is very different from catching someone who simply happens to be scrolling through a feed.
Lower Funnel? Absolutely.
Would I consider ChatGPT advertising a lower-funnel tactic? Absolutely.
Not because every ChatGPT conversation ends in a purchase, but because many users arrive already looking for an answer, asking questions, comparing alternatives or trying to decide what to do next. That places the advertising opportunity much closer to the decision stage.
Someone watching television might notice your brand. Someone seeing your display ad might remember it. Someone asking AI to help compare options may already be several steps further along.
That does not mean ChatGPT belongs only at the very bottom of the funnel. It can certainly play a role in consideration too, but for the right client I see a particularly interesting opportunity between consideration and decision. And that is where this stops being just an interesting AI story and starts becoming a useful media-planning conversation.
So, Should It Be in Your Media Plan?
Maybe. And I mean that seriously.
The fact that a new tactic exists is never a good enough reason to buy it. The audience, objective, message and level of media investment all need to make sense. That is the part that often gets lost whenever a new advertising product launches.
New does not automatically mean useful. Useful means it solves something.
Maybe your client needs another lower-funnel layer. Maybe paid search is already doing a lot of heavy lifting and you want another environment where intent is present. Maybe your audience is actively using AI to compare options before making a purchase. Or maybe none of those things are true. That is why strategy still has to come first.
What I Would Be Watching
If I were sitting on the client side deciding whether this belongs in the plan, I would be watching more than impressions or clicks. I would want to understand what kinds of conversations create opportunities for the brand to appear, whether traffic behaves differently from traditional search or display traffic, and which products or services naturally fit this environment.
I would also want to see how ChatGPT works alongside the channels already doing the job. Because this is not about replacing Google, social or any other established channel. It is about adding another tool when that tool has a clear role.
Where We Go From Here
For Dalilu Communications, ChatGPT advertising is no longer something we are simply observing. We are ready to start incorporating it into media plans where the audience, objective, message and media investment make sense, particularly as a lower-funnel tactic for clients who want to reach people while they are asking questions, comparing options and moving closer to a decision.
I expect we will learn a lot as we start using it, and that is part of being early. You do not pretend to know everything about a new channel before the market has had time to mature. You understand what it can do, place it carefully, measure it properly and learn from what happens.
If you are a marketing director reading this and wondering whether ChatGPT advertising has a place in your own media mix, that is probably the right question.
Not: “Should we be doing AI?”
But: “Is our audience already using AI at a point where our brand could genuinely be useful?”
If the answer is yes, there may now be another place worth showing up. And if you would rather have someone else figure out where that place belongs in the media plan, well… that happens to be what we do.
Sources
Additional sources for new section: Google Gemini (Aug. 2026); OpenAI Ads / ChatGPT Ads availability (Sep. 2026); Anthropic on Claude ad-free positioning.
OpenAI — Expanding access to AI with ChatGPT ads — https://openai.com/index/expanding-access-to-ai-with-chatgpt-ads/
OpenAI Help Center — Ads in ChatGPT — https://help.openai.com/en/articles/20001047-ads-in-chatgpt
OpenAI Help Center — Ads Manager Availability — https://help.openai.com/en/articles/20001245-ads-manager-availability
Pew Research Center — Americans and AI 2026 — https://www.pewresearch.org/internet/2026/06/17/americans-and-ai-2026-chatbots-smart-devices-and-views-on-impact/
by Michaela Sharp | Sep 23, 2026 | Blog, Dalilu Wire, Digital, Media Buying, Media Planning
A reflection on programmatic advertising, hands-on media buying, and the learning mindset behind staying ahead of the curve.
Lately, I feel like I have been on one of those hockey-stick learning curves.
A very Canadian analogy, I know. You know the shape I mean. For a while the line barely moves. You are learning, experimenting, getting things wrong, trying again and wondering whether you are actually getting anywhere. Then something clicks. Suddenly the line shoots upward. I have been experiencing that again recently while learning video editing and experimenting with completely new ways of presenting business content. And this morning, Grant said something about us, as business owners, always trying to stay ahead of the curve. It made me think about how many times we have actually done this. One of the biggest examples for us was programmatic advertising.
And before I go any further, because I realize not everybody spends their days thinking about media buying, let me explain what I mean.Programmatic advertising is essentially a way of buying digital advertising through technology rather than negotiating every individual placement directly with a website or publisher. It allows advertisers to reach specific audiences across thousands of websites, apps, streaming environments and other digital spaces, often in real time. If you have ever been reading an article online and noticed a banner ad, watched a video with an ad before or during it, or used an app where advertising appeared, there is a good chance some of that advertising was purchased programmatically. That is the simple explanation.
The complicated part is how it is bought, where those ads actually appear, who is seeing them and whether anyone is paying enough attention to what happens after the campaign is turned on. And that is really what this story is about. Because when programmatic advertising first entered my world, I was fascinated by it. I was also suspicious of it.
My Relationship With Programmatic Advertising Started With Suspicion
When programmatic advertising first started becoming part of the media world, I was still working at an agency as a media director. At that time, programmatic felt a little like a newborn trying to hold its head up.
Interesting. Full of potential. But still very wobbly.
As somebody who has always liked being hands-on with my media buys, I immediately had mixed feelings about it.
I loved the idea.
What I didn’t love was handing a major part of a client’s campaign to a third party and then waiting for somebody else to tell me what was happening. I wanted to understand it. More importantly, I wanted to run it myself. A couple of years later, Grant and I started Dalilu Communications, and suddenly that same question came back – only now it mattered even more. These were our clients. Our recommendations. Our reputation.
Then we were approached to work on a national campaign where programmatic advertising needed to be part of the media mix. We contracted two different vendors to run portions of the campaign. And that experience changed everything for me.
The Numbers Looked Fantastic. That Was the Problem.
On paper, the campaigns looked great. High click-through rates. Low CPMs. Lots of impressions. If you looked only at the report, you could easily conclude that the campaign was performing beautifully. But then I started looking at where the advertising was actually appearing and where some of the traffic was coming from. Back then, you could see much more of this information inside Google Analytics.
Oh, back in the days, how I loved that feature.
Today?
Different story.
Now we deal with privacy changes, modelling, estimates and much less visibility into some of the information we used to be able to dig into directly. But back to the story. When I looked at some of the placements and traffic patterns, I was appalled. Some of them simply didn’t make sense. So naturally, we started asking questions. Why were we appearing here? Why was this domain generating so much traffic? What was being optimized? What exactly were we buying?
The more questions we asked, the slower the answers became. Sometimes we received answers that technically contained words but somehow still didn’t answer the question. It felt a little like talking to customer service at a large department store. You know there is probably a logical explanation somewhere in the building, but nobody you are talking to seems able to give it to you. Eventually Grant and I looked at each other and basically said:
We can do this better.
And if we were going to recommend programmatic advertising to our clients, we wanted to understand exactly what we were recommending.
So We Learned How to Do It Ourselves
We found a programmatic platform headquartered in Chicago with Canadian connections and immediately started training. Within about a month, we were on a plane to Chicago to complete their certification program and become programmatic tactic certified professionals.
At the time, our children were still very young. Our youngest was only two. For me as a mom, leaving for four days was difficult. My parents don’t live in Canada, and we didn’t have grandparents available to simply move into the house for a few days and take over. We had to arrange childcare and trust that everything would be okay while we were gone. I remember being terrified about leaving. Professionally, however, it turned out to be one of the best decisions we ever made. Learning to run programmatic campaigns ourselves gave me two things I value enormously:
Freedom and confidence.
And something else happened that I don’t think I fully appreciated at the time. Running the campaigns myself began making me a better media planner.
Hands-On Learning Changes the Way You Think
There is something completely different about learning a media platform by actually working inside it every day.
You learn by doing. You learn what happens when you change a frequency cap. You see which domains behave strangely. You discover which audiences respond differently. You learn when a targeting strategy is too narrow and when it is far too loose. And you learn things about audiences that no textbook, presentation or vendor sales deck can really teach you.
Retargeting campaigns are a great example. By watching an audience over time, I can start seeing patterns in the types of websites and content that audience consumes. That knowledge doesn’t disappear when the campaign ends. Maybe months later I am planning a completely different campaign for another client trying to reach a similar audience. Now I have somewhere to start. I have learned something. Maybe I have discovered a group of domains frequently visited by that type of audience. Nobody handed me that list. Nobody sent me a PowerPoint saying, ‘Michaela, here are the websites you should buy.
I learned it through planning campaigns, buying the media, watching the behaviour and optimizing the campaigns myself. Then I carry that knowledge into the next strategy. That is where being both the media planner and the media buyer becomes incredibly valuable. One part of my work continuously teaches the other. Perhaps that is the VIP treatment our campaigns receive. Not because there is some secret formula available only to us.
It is simply because the person who developed the strategy is also looking at what is actually happening inside the campaign.
Creative and Media Cannot Live in Separate Rooms
Hands-on buying also changes the way you think about creative. Creative and media cannot live in two separate rooms. If a piece of creative is trying to communicate a lot of information, for example, perhaps the audience needs more opportunities to see it. A higher frequency might make sense. If the message is extremely simple, the answer might be completely different. There isn’t one magical frequency number that works for everything.
It depends.
What are we asking people to understand?
How complicated is the message?
How long is the campaign?
How large is the audience?
What other media are supporting it?
How many creative executions are running?
There are always variables. Because I both plan the media and operate campaigns, what I learn from one campaign feeds directly into the next media plan. That is probably the part I value most.
Planning defines what the campaign should accomplish.
Buying shows me what actually happens.
Optimization helps me understand why.
And the next plan becomes better because of everything I learned from the campaign before it.
Sometimes a Great CTR Is a Warning Sign
This is also why I don’t automatically celebrate when I see an unusually high click-through rate. Someone might look at the report and think: Fantastic! Look how many people clicked.
My reaction is often: Why?
And then I start investigating.
If I am running page-load tracking on the campaign landing page, I can compare ad clicks with the people who actually reached the website.
Where are those clicks coming from?
Are people really reaching the page?
Is one domain generating an unusually large percentage of the activity?
Does the behaviour actually make sense?
Sometimes a placement produces numbers that are simply too good to be true. And instead of celebrating it, I pause it.
Because great numbers don’t always mean great performance.
Sometimes they are a red flag. Imagine running a province-wide campaign and suddenly noticing an unusually high level of engagement coming from one tiny rural community in the middle of nowhere. Maybe the town has spectacular internet. Maybe everybody there happens to be incredibly enthusiastic about whatever we are advertising. Maybe there is a perfectly logical explanation. But my investigator’s instinct is still going to kick in. If a community of a few thousand people suddenly appears to be generating hundreds and hundreds of clicks, I want to know why.
There is something fishy in the data. Could it be legitimate? Sure. But that is exactly when the little red light in your head should come on. Don’t celebrate yet. Look.
Programmatic Impressions Are Not Garbage
Every once in a while, I hear someone say that programmatic impressions are garbage. Actually, let me be fair. I have heard one particularly loud voice saying it repeatedly. And I strongly disagree. Social media is an incredibly powerful advertising tool. We spend a ridiculous amount of time scrolling through social platforms and living inside our comfortable, beautifully curated little bubbles. But surely we still do other things online. We read articles, check the news, read entertainment stories, research things, visit websites, stream video, and yes, some of us play cute little games on our phones when nobody is watching.
So no.
Programmatic advertising is not garbage.
Badly managed programmatic advertising can absolutely be garbage.
There is a very important difference. If programmatic isn’t working well, perhaps the question shouldn’t automatically be whether programmatic itself is useless.
The first question should be:
How is it being used?
Media Works Better When It Works Together
There is another reason I don’t believe in dismissing programmatic in favour of social media.
Reach.
If your entire digital strategy exists on one or two social platforms, you are intentionally limiting your brand to those environments. You might reach a lot of people there. But those platforms still represent only one part of their overall media day. Someone might see your brand on social media in the morning. Later they read an article and encounter the campaign again. That evening they stream something. Perhaps the next morning they hear your radio commercial while driving. Then two days later they Google the product.
Which impression caused the conversion? Good luck answering that perfectly.
That is why I have always believed in looking at media holistically rather than constantly asking which single platform is the winner. The media can work together. Expanding across appropriate tactics can increase reach, build familiarity and may create a greater opportunity for conversion. You’ll notice I deliberately say may or can. I don’t say it will convert. Because media isn’t black and white. There are always variables.
The offer. The product. Creative. Competition. Timing. Budget. The website. The economy. The audience.
And probably another twenty things we haven’t even thought about.
That uncertainty is part of what keeps this job challenging.
And, most of the time, fun.
The Problem Is Often “Set It and Forget It”
If a programmatic campaign is launched with extremely loose targeting, no thoughtful domain strategy, very little optimization and nobody regularly reviewing where the budget is actually going, you can burn through an enormous number of impressions while accomplishing very little. Mobile games are a good example.
There is nothing inherently wrong with advertising inside a game if the audience and campaign make sense. But if an enormous percentage of your daily budget starts disappearing into gaming apps, questionable websites or placements producing suspicious amounts of activity, somebody should ask why. And this isn’t only a programmatic problem. I see the same mentality with Google Ads and other digital platforms.
Set it up. Turn it on. Come back later. Generate the report. Look at the numbers. Done.
Digital media shouldn’t be a set-it-and-forget-it product.
Someone needs to be looking under the hood.
Ask What You Are Actually Buying
This is one of the simplest things I wish more marketers and clients would do. Ask questions before committing to the buy. Not ridiculous questions. Not questions for the sake of proving that you know marketing terminology. Basic media-buying questions.
What CPM are we buying at?
How many impressions are included?
What frequency should we expect?
Where can the advertising appear?
What exactly are we optimizing toward?
Then, once the campaign is running, ask to see placement or domain reporting when it is available and relevant. Look at it. You don’t need to panic every time you see a website you have never heard of. We all visit strange corners of the internet. There will always be noise. Nobody is going to create a perfectly pristine internet. But patterns matter. CPM in particular still surprises me. We work in a relatively small market, and I regularly see proposals where people focus almost entirely on the total price without asking one of the most basic questions before committing: What am I actually getting for that money?
If someone tells you a digital campaign costs $10,000, the $10,000 alone tells you very little. The proposal will usually tell you the audience. It will usually tell you the geography. It will tell you the campaign dates. Those things are expected. But what about the actual media inventory? How many impressions does that investment buy? What is the CPM? That matters. A proper media plan should make those numbers clear enough that you can understand what is actually being purchased.
Ask Questions. But Don’t Destroy a Good Vendor.
There is another side to this conversation that I think is just as important. Ask questions.
Absolutely.
Hold your media partners accountable. But don’t make their lives miserable simply to prove that you are paying attention. A good vendor is incredibly valuable. If somebody is proactive, transparent, thoughtful and genuinely looking after your campaign, treat that person like the professional they are. And I say this knowing full well that I am one of those clients who asks a lot of questions. I know I can be tough. Probably very tough sometimes. And I am still learning to find the balance I am talking about.
But after doing this for as long as I have, I can usually tell the difference between somebody who has made a reasonable decision I happen to question and somebody who is giving me an answer because they haven’t actually been paying attention to the campaign. You see it in the numbers. You see it in the answers. You see it in how quickly someone understands what you are asking.
And perhaps this brings the whole story back to why Grant and I decided to bring programmatic buying in-house in the first place. We believed we could do a better job for our clients. There was also another very practical incentive. We own the company. If we don’t do good work, clients eventually don’t hire us. And if clients don’t hire us, there isn’t magically a paycheque arriving for Dalilu at the end of the month. That creates a particular kind of motivation. I am not saying that someone working for a salary doesn’t care about doing excellent work.
Many of the best people I work with are employees of other companies and care tremendously. But ownership does change the equation. When your name, reputation and livelihood are tied directly to the work, you feel it differently. And maybe that is another reason I have always wanted to understand what is happening inside the campaigns rather than simply accepting the report. Still, balance matters. I have vendors I trust enormously because over time they have shown me that they care about our clients’ campaigns. Those relationships matter too.
Good media has always required judgement. Technology hasn’t changed that. If anything, technology has made judgement even more important.
Being Ahead of the Curve Doesn’t Mean Knowing Everything First
Thinking about all of this now, I don’t believe Grant and I have somehow been able to predict where advertising or technology is going. That isn’t what being ahead of the curve means to me. It means noticing that something is changing and becoming curious enough to start learning before you are forced to. Programmatic was one of those moments. Today, I find myself doing it again with video. AI is another obvious one. And I am sure there will be another one after that. The tools keep changing. The uncomfortable learning periods keep coming. And every time, at the beginning, it feels like I know absolutely nothing. The line barely moves.
I make mistakes. I Google things. I watch tutorials. I try something. I hate it. I redo it.
And then, without really noticing exactly when it happened, things start connecting.
The curve changes.
That might actually be the biggest lesson I have learned from trying to stay ahead of anything. You don’t need to know everything first. You just need to be willing to start learning while the line is still flat. Because if you keep going long enough, sometimes that learning curve starts to look very Canadian.
Like a hockey stick.
