AI Will Help Advertisers and Shoppers

In what seems like no time at all, artificial intelligence has become a mainstream business tool. One of its most promising uses is in advertising.

AI will likely improve advertising efficiency, boosting key metrics such as return on advertising spend and conversions while engaging shoppers better.

Here’s how.

Personalization

In a recent Harvard Business Review article, academics from Babson College, the University of New Hampshire, and the University of South Carolina proposed a framework for how businesses should employ two forms of artificial intelligence for marketing.

The article distinguishes generative AI (ChatGPT, others) from analytical (data analysis), arguing that the latter helps predict shopper behavior and outcomes, while genAI is creative.

This tandem of AI tools will facilitate personalized ad messaging to individual shoppers or micro-segments at scale. It is the idea of precise ad targeting combined with personalization to drive conversions.

Romain Lerallut

“It’s no longer just about reaching a large audience; it’s about reaching the right audience with tailored, relevant messaging that delivers better outcomes,” wrote Romain Lerallut, head of Criteo AI Lab, in an email conversation with Practical Ecommerce.

Analytical AI will create target audiences and segments, while generative AI will deliver the just-right message, ultimately improving ad performance.

Targeted personalization is not easy, even for data scientists, but most advertisers will not have to manage it directly. Rather, advertising technology companies — think ad exchanges and ad services — will likely offer personalization as a service.

Budget Allocation

Allocating advertising budgets is instinct-driven and necessarily relies on human experience and data silos.

Most marketers can interpret relevant data and make correct budget decisions. Yet automated budget allocation in a fast-moving digital ad marketplace would be an improvement.

AI will improve budget management, which, in a sense, builds on the targeted personalization described above.

AI models will analyze performance, customer behavior, and real-time market signals to optimize budgets across channels and campaigns.

“Brands can target consumers more precisely and make every advertising dollar count,” wrote Criteo’s Lerallut.

Most advertising platforms will likely include predictive budget allocation tools that automatically shift ad spending toward the top-performing audience segments or channels. Advertisers will benefit from improved performance.

Shopper Journey

AI can group disjointed data sources to better understand how shoppers become customers.

In practice, a predictive AI budget allocation tool will not rely on last-click attribution. Instead, analytical AI and machine learning tools will power marketing mix models and multi-touch attribution models.

Per Lerallut, “As consumers engage with brands across more touchpoints than ever — online, in-store, on social media, through apps, and more — AI can help unify and interpret these interactions to create a cohesive view of the customer.”

Advertisers will likely require marketing mix models and multi-touch attribution tools. Google and Meta currently offer free solutions.

Frequency Optimization

With a clear map of the shopper’s journey, AI can improve how and where ads appear, optimizing placement, frequency, and timing.

This could also mean optimization beyond a single platform. AI could suggest frequency caps on Meta, Google, and other platforms to avoid reader fatigue.

Perception

Combined, AI’s capabilities will result in a better shopper experience. Personalization, budget allocation, customer journey, and ad timing will lead to a deeper understanding and trust with the advertiser’s business.

Ultimately, AI will help both advertisers and shoppers.

The Rebirth of ‘Marketing Mix Modeling’

Lingering privacy challenges and ever-improving cloud and artificial intelligence technology are driving a marketing model renaissance.

Marketing mix modeling (MMM), launched in 1949, fell out of favor in the early 2000s when digital advertising took off. The data-driven technique had long helped marketers understand how variables such as advertising, promotion, and prices impact revenue.

Yet compared to tracking cookies and last-touch attribution models, MMM seemed complex and expensive.

Renaissance

In 2025, however, MMM has enjoyed renewed attention.

Meta and Google have released free, open-source MMM tools in the past couple of years — Google’s Meridian on January 29, 2025, and Meta’s Robyn in 2023.

Why does MMM interest two of the largest digital advertising platforms? I see three probable factors: tracking cookies, AI, and cloud computing.

Google launched Meridian, an open-source marketing-mix model, last month.

Cookie-less Advertising

Controversies surrounding tracking cookies are the first driver. Cookies are a foundational and useful technology. A first-party cookie on a browser keeps users logged into a website and retains their preferences.

However, third-party tracking cookies that catalog an individual’s behavior across web properties are a privacy pariah. Laws such as Europe’s General Data Protection Regulation and the California Consumer Privacy Act limit such cookies, and many browser companies have stopped supporting them entirely.

The potential of cookie-less ad targeting makes MMM attractive to large-scale advertisers and platforms.

Advertising performance. Third-party cookies, despite privacy concerns, drive ad targeting and thus performance. MMM should help advertisers identify which marketing channels and creatives produce the best returns. Coupled with new ad targeting techniques, MMM will almost certainly improve performance.

Meta’s Robyn, for example, helps advertisers analyze the performance of campaigns across Facebook, Instagram, and other channels. It gauges channel effectiveness and optimizes ad spend based on results.

The era of cookie-less targeting encourages the use of MMM. Some of the most forward-looking, high-budget advertisers are considering alternative targeting methods and new promotional channels. Monitoring those experiments requires complicated multi-touch attribution or MMM.

For example, Google’s Meridian MMM moved beyond standard regression models to a theory called “Bayesian causal inference,” which captures the impact of imprecise marketing actions, such as a social media post.

Personal privacy is yet another reason why MMM appeals to Google, Meta, and many advertisers. The model aggregates data and generally avoids personally identifiable information.

Artificial Intelligence

AI makes MMM relatively faster, more adaptive, and easier to scale.

Improved speeds come first from training the model quickly. The foundational, now-available models are a massive headstart compared to starting from scratch.

Second, AI helps process and clean large, complex datasets from multiple sources such as digital ads, TV, print, and online and in-store sales. The associated algorithms detect seasonality, outliers, and data anomalies, reducing manual work but still requiring data scientists to fine-tune the models.

Regardless, the AI behind Meta’s Robyn dynamically adjusts model variables, improving accuracy automatically. It is thus more adaptable and scalable.

Cloud Computing

Twenty years ago, the rise of web-based marketing produced massive datasets and hefty processing loads. Analyzing that info typically required custom-built infrastructure and expensive data warehouses

These limitations no longer exist thanks to cloud computing advances and affordability. Instead of spending $500,000 or more on MMM software and servers, a company can run Google Meridian in the cloud for a fraction of the amount, perhaps as little as $10,000 a year.

Accurate modeling, however, requires some scale — businesses investing at least $500,000 per year in advertising likely benefit the most. But that could change if MMM becomes available as a service.

The New Era of Cookie-less Ad Targeting

Advertising technology has entered a new phase that pairs personal privacy with targeting.

For years, the ad industry has depended on third-party tracking cookies — tiny snippets of code downloaded to a user’s browser — to identify prospects, gather data about them, and employ that data for targeting. Consumers saw relevant ads, and advertisers enjoyed good, if not superb, returns.

Adobe’s new real-time customer data platform is one of many innovations in cookie-less ad targeting.

Cookies

Unfortunately, cookie-based tracking has privacy problems. In contrast to first-party cookies, which store preferences, third-party tracking cookies aggregate, share, and thereby expose lots of private information.

Just this month, Wired magazine reported, “Google’s advertising ecosystem reveals that a wealth of sensitive information is being openly served up to some of the world’s largest brands…Experts say that when combined with other data, this information could be used to identify and target specific individuals.”

Privacy advocates recognized this and began railing against sharing personal and sensitive information across advertising networks. Technology companies responded. Mozilla’s Firefox browser, for example, stopped allowing tracking cookies back in 2019.

Google, which operates one of the largest ad networks in the world and has the most popular web browser in Chrome, nearly killed third-party cookies in 2024, as it planned to block them in Chrome and move all advertisers to alternative targeting methods — but it didn’t happen.

“Google heard the alarm bells from the industry that nearly 20 years of ad tech infrastructure cannot be recreated in six months,” said David Stein, then the CEO of the data firm Audigent, in an April 2024 email to Practical Ecommerce.

Those alarm bells could have been Google’s ad exchange customers, its ad tech partners, and the anti-trust regulators concerned that ending cookie tracking would give Google a performance advertising monopoly.

Hence tracking cookies remained in Chrome, but ad tech providers did not stop innovating.

Innovation

The same day that Wired published its report, Adobe announced the general availability of its real-time customer data platform (CDP).

“Brands have long relied on third-party audience signals to power tailored digital ads. As consumers play a more active role in customizing their privacy preferences, a move away from third-party data means new tools are required for brands to identify relevant audiences and deliver personalized ad experiences,” read the Adobe press release.

Adobe’s CDP allows advertisers and publishers to securely collaborate on first-party data to enhance ad targeting while respecting user privacy. This relationship is often called second-party data sharing, and resembles Shopify’s Audiences and similar products.

Methods

Second-party data sharing is one of six primary methods ad tech companies are exploring as they try to maintain or improve ad targeting efficacy without raising privacy concerns.

• Second-party data sharing allows advertisers and publishers to collaborate, often sharing aggregate data.

• First-party data is when a business uses its own data to target or retarget individuals or cohorts. Many companies are buying publications to advertise to known customers and prospects.

• Unified ID solutions produce encrypted identifiers to share among systems. The Trade Desk’s Unified Solution is an example. Often, these IDs stem from an email address.

• Data clean rooms make it possible for analysts to use data without privacy concerns.

• Cohort-based advertising targets groups of shoppers with similar profiles instead of individuals.

• Contextual targeting analyzes the context — web page, app, video, email — to deliver relevant ads to consumers without relying on personal data. Artificial intelligence makes this more successful.

Used individually or in combination, these targeting techniques have significant promise. Most reports suggest that results vary among advertisers, yet the techniques will likely work better than third-party cookies.

Ad Experiments

Listing the ad tech companies developing cookie-less targeting methods would fill pages. Thus advertisers seeking to acquire customers might find it wise to experiment.

For example, marketing platform Zeta Global recently purchased the email advertising company LiveIntent. Some in the industry believe Zeta Global valued LiveIntent’s ability to identify website visitors based on email interactions.

Could Zeta Global’s demand-side platform be worth a look?

Similarly, Paved, another email ad platform, expanded its programmatic network, permitting advertisers to target and retarget shoppers safely and privately.

In short, ad targeting is changing, and the best options may not be familiar platforms.

Print Ads Are an Ecommerce Opportunity

Print magazines can be an unusual opportunity for niche ecommerce merchants.

Digital advertising dominates ecommerce, with good reason. The ads are measurable, easily updated, and often produce instant results. Place an ad and get a sale on the same day.

In contrast, print advertising does not usually produce immediate results and is not easy to measure. Yet the medium has unique values that established marketers appreciate.

Take Amazon’s October 2024 printed toy catalog, for example.

“We’re not surprised to see Amazon leaning into their toy catalog,” said Polly Wong, president of Belardi Wong, a New Jersey-based marketing firm, in a correspondence with Practical Ecommerce.

“Hundreds of brands have added direct mail and catalogs to their mix over the last five years because print offers unique advantages to digital marketing — more real estate to show your product and entice consumers to buy, 100% reach, high lifetime value from customers who respond to print, and a very effective way to stand out to new customers.”

In essence, print promotions — such as catalogs or magazine ads — complement digital advertising.

Brand Advertising

Magazines and printed catalogs can provide valuable brand advertising, an often overlooked approach by ecommerce marketers.

Brand ads in a magazine tend to linger and get many “impressions.”

“An ad in print will sit in someone’s home or near someone’s desk for months upon months, sometimes even years,” wrote Troy Klongerbo, general manager at Homestead Living magazine, in an email interview.

To Klongerbo’s point, most magazine subscribers flip through the publication several times, repeatedly referring to articles and ads. His publication often includes recipes and reference material so that issues become “fixtures” in the kitchen or the shed.

Print magazines such as Homestead Living can linger on a coffee table or desk for months as subscribers revisit them.

Printed ads can also build trust. A company with a printed ad appears to be established and trustworthy.

“There’s a deeper implied trust with words and pictures placed in print. Once printed, words and messages cannot be altered, providing for a sort of permanence. Digital ads, even the best of them, are ephemeral to a degree,” wrote Klongerbo.

Indications

While short-lived, digital ads provide solid data, which print does not. An advertiser must be clever about collecting and analyzing print performance and willing to accept indications.

Many print advertisers use several performance indicators in combination.

• Brand lift surveys. Measure changes in brand awareness, consideration, and favorability after exposure to brand advertising campaigns.

• Social media engagement. Monitor mentions, shares, and sentiment related to the brand, especially when an issue is first published.

• Website traffic. Pay attention to website traffic and engagement metrics. An increase in direct or branded search visits often indicates that print advertising is working.

• Share of voice. Compares the brand’s advertising or marketing presence against competitors within a specific market or industry.

• Direct response. Finally, thanks to coupon codes, subdomains, and QR codes, magazines and catalogs can capture some direct response data, but this, too, is indicative, not definitive. Ten scanned QR code visits may represent 100 consumers.

QR and coupon codes enable print ads to capture direct response data, as in this example from Homestead Living.

Affiliates

Some print publications use this last form of measurement — direct response — to run affiliate ads for ecommerce brands.

This affiliate relationship is relatively new in print magazines, perhaps borne out of influencer marketing. Modern advertisers are often comfortable with influencer relationships combining flat-rate and affiliate payments.

For example, a direct-to-consumer brand might pay an influencer a nominal fee to produce an Instagram Reel and then pay affiliate commissions to the influencer for sales made with her discount code.

Similar arrangments now occur in print. The advertiser pays a low flat rate to help offset the cost of printing and distributing the ad and shares the revenue tracked via a QR code.

If a brand’s typical affiliate share is 15%, the publication might receive 30% to 50%, recognizing that some shoppers will search for the brand or type in its URL rather than using the QR code.

Complementary

Branded print advertising can positively impact digital. Countless published reports confirm that branding boosts performance ads, driving relatively more clicks, engagements, and conversions.

In short, printed and online ads are complementary approaches to achieving marketing objectives.

Outstanding Marketing Campaigns from 2024

Every December we publish a rundown of innovative marketing campaigns for the year. We handpick campaigns from established companies that have received awards and recognition or that we encountered directly.

Here is a list of outstanding marketing campaigns for 2024. The campaigns feature collaborations, dupes, celebrity spots, mobile apps, mascots, dance routines, pop-ups, and AI-powered videos.

Heineken – Boring Mode

Heineken – The Boring Phone

In a campaign to help people socialize, Heineken released The Boring Mode app. With a single swipe, The Boring Mode turns a smartphone into a boring phone, blocking other social and work apps, notifications, and even the camera for a set period. During events, including the Amsterdam Dance Event and Live Out Festival in Mexico, Heineken activated hidden messages when people held up their mobile phones to encourage them to turn their smartphones boring and keep the moment in their memory, not on their phones.

–

McDonald’s – McDonnell’s Chicken Big Mac Pop-Up

McDonald’s Chicken Big Mac

McDonald’s secretly introduced a Chicken Big Mac, a spin on the classic menu item through a dupe of an L.A. pop-up called McDonnell’s by Chain. McDonald’s also enlisted influencer Kai Cenat to stir debate on whether or not a Chicken Big Mac is a Big Mac. McDonald’s then launched the Chicken Big Mac on October 10, allowing followers to order the item and decide for themselves.

–

Squarespace – Rick Rubin

Squarespace teamed up with creative visionary Rick Rubin, co-founder of Def Jam Recordings and former co-president of Columbia Records, to build Tetragrammaton, an online world of curated materials, and a new website template, Transmission. Inspired by Rubin’s Tetragrammaton, Transmission is a minimally designed, customizable framework for anyone to share their interests, with layouts showcasing audiovisual media and members-only content.

–

Qatar Airways – Star in Your Own Adventure

Qatar Airways launched an AI-powered “Star in Your Own Adventure” campaign, in which viewers select from multiple scenes within a film to star in the leading roles. Through AI technology, the characters become a reflection of the viewer’s appearance, adapting to their facial features and skin tone. The immersive experience now offers holiday-centric moments, including an elf, snowman, gingerbread person, and cheerful characters, inviting participants to star in their personalized holiday-themed Qatar Airways film.

–

Calvin Klein – Jeremy Allen White

Calvin Klein has launched its winter 2024 collection campaign, starring Jeremy Allen White from “The Bear” television series. This is the third global campaign with White. While the previous two focused on White in Calvin Klein’s emblematic underwear, the new campaign features classic, understated styles, demonstrating the brand’s ability to remain relevant and give fans what they want.

–

Liquid Death – E.l.f. Cosmetics

Liquid Death canned water collaborated with E.l.f. Cosmetics to launch Corpse Paint. Liquid Death wants to help consumers “murder” their thirst and prefers irreverent and organic marketing over advertising. Brand collaborations allow Liquid Death to inject its spirit and provide the partner brand with an opportunity to expand and do something unexpected. The Corpse Paint collaboration captured more than 250 million social impressions. Other recent collaborations with Liquid Death include sparkling water with ice cream chain Van Leeuwen and a casket cooler with Yeti.

–

Airbnb – Polly Pocket

Mattel is celebrating Polly Pocket‘s 35th birthday by making the pocket-sized doll an Airbnb host and opening her two-story Slumber Party Fun compact, a 42-foot-tall structure in Littleton, Massachusetts. Guests experienced a retro picnic, beauty fun with Polly’s vanity, fashion try-ons with Polly’s closet, slumber party fun in Polly’s Action Park Tent, and more. Guests booked stays for September 12 to 14. Polly’s home was also opened for 21 daytime experiences for up to 12 guests each from September 16 to October 6. Guests were responsible for travel to and from Littleton, Massachusetts. The location showed the popularity of Airbnb.

–

Gap – “Get Loose” with Troye Sivan

Gap ran a fall 2024 campaign called “Get Loose,” which starred global pop star Troye Sivan and the dance group CDK Company. Set to Thundercat’s viral hit “Funny Thing,” the film features original choreography showcasing dancers in baggy and loose-fit denim. Sivan is known for his music videos, featuring elaborate and sometimes racy dance routines, and has a combined social following of over 39 million. As part of the campaign, Sivan created an in-store playlist on Spotify. With the Thundercat soundtrack, Gap is tapping into viral TikTok songs, celebrity ambassadors, and dance-heavy advertising to reach Gen Z consumers.

–

Pop-Tarts – Edible Mascots

In 2023, Pop-Tarts signed a multi-year agreement to become the title sponsor of the college football Pop-Tarts Bowl game, originally commissioned as the Sunshine Classic. To celebrate the event, Pop-Tarts launched the “first edible mascot” last December. The Kansas State football team devoured a Frosted Strawberry mascot in a viral event on social media. For this year, the game’s MVP will select one of three flavors to “ascend to ‘mouth heaven.’” In the run-up to the game, consumers can guess the third flavor, purchase limited-edition Pop-Tarts Printed Fun Mascot Packs, and follow along on Instagram and TikTok.​​

–

Coca-Cola – AI-powered Christmas Ads

Coca-Cola launched three short AI-generated video ads for Christmas, and the results have sparked backlash on social media and in the press. Three AI studios (Secret Level, Silverside AI, and Wild Card) produced the ads using the generative AI models Leonardo, Luma, Runway, and Kling. The ads highlight the weaknesses and limitations of current video-generation models, such as difficulties with proportionality and natural movement. While the ads are essentially derivatives of the Coca-Cola Christmas ads from 1995, the new AI-powered experiments have drawn much attention and sparked conversation, which is not necessarily bad.

Attributing Marketing Expenses by Channel

Some marketers now measure and attribute the cost of labor, technology, and services to individual promotional channels.

The concept is simple. Many companies track only revenue from marketing channels without considering the expense of managing them. The result is often misleading bottom-line performance.

Channel Comparison

Imagine a business with two marketing channels, A and B, each costing $1,000. Both generate 3,000 interactions from potential customers. However, Channel A converts at 2.5%, while Channel B converts at 4%.

If both channels had a $75 average order value and a 25% gross profit margin, Channel A would produce $406 in profit, and Channel B would earn $1,250. Channel B is the clear winner when compared in this way.

Just about every business would take the $1,000 invested in Channel A and double down on Channel B. After all, Channel B produces about three times as much profit.

This is often the proper choice, but not always.

Marketing Budgets

There’s more to marketing expenses than advertising or accessing a channel.

There are salaries for the marketing team, software subscriptions, creative design expenses, and even influencer fees.

Let’s apply this idea to Channel A and Channel B. Suppose each channel is a demand-side platform (DSP), wherein marketers choose from a list of potential publishers.

DSP A allows marketers to pick some basic targeting demographics, but there’s little a specialist could do to optimize performance. It is a set-it-and-forget-it sort of platform.

On the other hand, DSP B has 100 targeting options that can be compared, fine-tuned, and optimized.

DSP B’s platform provides real-time data with Slack notifications every time a campaign’s conversion rate changes.

The marketing specialist spends about 30 minutes a month setting up the simplistic DSP A but about an hour a day monitoring, studying, and tweaking DSP B.

If the marketing specialist earns $50 an hour, DSP A costs about $25 per month in labor. Given 20 working days a month and an hour per day spent monitoring and optimizing, DSP B takes $1,000 in labor to run.

When counting labor, DSP A generates $381 in profit compared to DSP B’s $250. DSP A is the clear winner.

Applying the Concept

Beyond labor, other expenses — e.g., software, creative design, agency retainers — could also change a channel’s return on investment, although not every expense is ongoing. Some are one-time or upfront charges that go away.

Thus, when attributing marketing expenses by channel:

• Decide what to measure. Labor, software, or simply the cost of an ad or promotion?

• Choose when to measure. Should the channel be measured per interaction? Or would monthly work better?

• Plan for upfront expenses. Should upfront expenses be amortized? If so, over what period? How will channels with amortized costs compare with those having ongoing expenses?

• Manage sensitive information. Some costs are sensitive or private. Will salaries be shared, or will the labor portion of the equation be closely held?

• Decide how you will measure. Should marketers use time-tracking software?

• Document the process. Record what, when, and how results are measured.

• Collect only essential data. There’s no need to track labor or software costs if they don’t impact marketing decisions.

Lastly, remember that sometimes the cure can be worse than the illness. Attributing expenses by channel can drive performance at the cost of damaging staff morale. So attribute with prudence.

Optimized Landing Pages Reduce Ad Costs

Optimizing a website for organic search rankings is well-known. Less common is optimizing for ad engines, crucial for managing the cost and effectiveness of paid traffic.

Ad targeting on social media and search engines relies on algorithms that bid for placements on advertisers’ behalf. Algorithms evaluate an ad and its linked landing page, assigning a relevancy score that influences the cost and performance. Meta calls the score “Quality Rank”; Google refers to it as “Ad Rank.” Both prioritize the relevance of an ad to the landing page.

Bidding engines on Meta, Google, and others optimize around a goal, typically conversions (for ecommerce). Low click rates on ads drive up their costs, as do landing experiences with low conversions.

But an optimized experience can reduce ad costs. Note the chart below showing my firm’s A/B test on Meta. The test traffic was split: half went to a standard ecommerce product page and half to an ad-engine-optimized version. Both used the same ad and budget. The optimized experience received 28% more traffic due to lower per-impression ad costs and drove 99% more revenue than the product page.

3 AEO Hacks

Ad engine optimization (AEO) prioritizes the relevance of a landing page to an ad. Little relevance produces high bounce rates and low conversions. Ad engines detect this and penalize the campaign with higher ad costs. Custom landing pages are the traditional fix. They work well for bottom-of-funnel and long-running campaigns but not for top-of-funnel customer acquisition, especially on social media.

For example, a shopper searching Google for a “red four-slot toaster” has a clear purchase intent and would likely convert from a landing page featuring red four-slot toasters.

But a consumer on Facebook might see an aspirational kitchen ad that happens to feature a red four-slot toaster, requiring a different landing approach.

Consider these AEO hacks:

• Mirror campaign creative. The simplest way to increase relevance is to mirror the images in the ad onto the landing page. This reassures shoppers they are in the right place, preventing an early bounce.

• Make it easy to buy all of the products. The top reason shoppers bounce is they can’t find the product, or it looks different. Incredibly, landing pages without the promoted products are common. It’s especially problematic on mobile, where the product may be swipes away from where the shopper lands. Hence ensure the landing page shows all products in the ad, not just the category or one item.

• Enable all intents. An ad showing a model on a beach wearing a t-shirt, shorts, and a hat could drive multiple interests: the t-shirt, shorts, summer, beachwear, or more. A landing experience focused on just one of those items will result in high bounces.

An optimized page makes it easy to buy, say, the advertised t-shirt while providing easy navigation to the other products. The page simultaneously shows all products in the ad and draws visitors into the broader category. The more visitors that browse, the more signals return to the ad engine, increasing the relevance score and lowering ad costs.

Algorithm Driven

To function, optimized pages require implementing the ad engine’s conversion API (“CAPI” for Meta and Google) to send the engagement signals. Testing and evolving the AEO experience is helpful — manually or automated.

Advertising on Meta, Google, and other channels is now algorithm-driven. Improving performance means optimizing for the algorithm. The combined quality of the ad and landing page directly impacts ad costs.