• Jocelyn Jiang

Ad Delivery Products Competitive Analysis

Updated: Nov 8, 2020

This article explores Facebook’s competitive positioning relative to other digital ad platforms such as Google, Amazon, and Microsoft. The note examines Facebook’s strengths and weaknesses across multiple focus areas relevant to ad delivery: delivery levers, automation, and performance guidance.

About the competitors

We analyzed each of the below competitors to understand their strategy and offerings for delivery levers, automation, and performance guidance. We chose this competitor set because they either are the top ad platforms in the US by ad revenue (Google, Microsoft, and Amazon) or provide consumer experiences that are most similar to Facebook’s (Twitter and TikTok).


As the industry leader with $105b in ad revenue in 2019, Google is projected to continue its market position in the years to come (with $128b 2022 ad revenue projected by eMarketer). Simultaneously, Google continues to diversify its revenue sources (e.g. going head-to-head with Amazon in commerce and cloud) and now sees just 84% of revenue from ads.

Advertisers use Google to reach the world’s largest search user base, the world’s largest video platform user base (YouTube), Google Display ads across mobile and web, and Google’s many proprietary destinations (e.g. GMail and Google Shopping).


Unlike other competitors in the digital ads space, Microsoft does not generate the majority of its revenue from advertising (with just 6% from ads compared to over 50% from cloud services). Still, with $8b in yearly ad revenue for FY 2020, Microsoft is 4th in ad revenue in the US (after Amazon) and 7th worldwide (after Baidu and Tencent). LinkedIn contributes around $2.5b in ad revenue.

Microsoft’s ad revenue comes primarily from Microsoft search products (e.g. Bing), LinkedIn, and other Microsoft properties. Advertisers frequently use Microsoft to attract a more mature and affluent audience, with 71% of the Microsoft Search Network’s audience as 35 and older.


After two years of >100% ad revenue growth, Amazon’s ad revenue growth slowed in 2019 and H1 2020. Still, the company is still seeing ~40% ad revenue growth YoY thanks in part to the surge of commerce dollars spent at Amazon. Amazon is now 3rd in US market share at 8.8% with $12b in revenue (which places it at 4th in worldwide ad revenue after Alibaba).

eCommerce advertisers use Amazon to place sponsored products ads, headline search ads, and product display ads. More recently, advertisers have described Amazon as a maturing ads platform and are settling into budget allocation strategies for sponsored products.


With $2.98b in worldwide revenue, Twitter ranks 6th in the US (after Verizon) and 9th worldwide (after Verizon, Baidu, and Tencent) in ad revenue. Still, the company is projected to grow slower than similarly sized competitors; eMarketer projects Twitter’s 2022 revenue to be just $3.2b.

Advertisers use Twitter to advertising via promoted tweets, promoted accounts, promoted trends, and other placements within Twitter. Despite less auction competition on Twitter than on Facebook, advertisers regularly report higher CPAs on Twitter.


A member of parent company ByteDance, China-based TikTok is the West’s hottest destination for short-form mobile videos. After generating just $177m in 2019 (a small fraction of ByteDance’s $17b revenue), TikTok is projected to earn $4.8b in ad revenue in 2020 – which would already place the app above Twitter and LinkedIn in ad revenue.

Advertisers use TikTok to reach the app’s nearly 1 billion users (that skew heavily towards the teen demographic) via feed-based video ads. Though TikTok advertisers report lower click-through rates than on competitor platforms, advertisers face less competition in ad auctions.

Competitive audit: Delivery levers

Delivery levers are the controls that ad platforms provide to help advertisers manage their campaigns. We analyzed each competitor’s offerings across the following delivery levers:

  • Bidding tools: How platforms enable advertisers to control cost per view or outcome

  • Budget management: How platforms empower advertisers to set and redistribute campaign budgets

  • Audience targeting: How platforms enable advertisers to target specific users or user groups

  • Campaign optimization: How platforms optimize delivery to achieve certain advertiser goals or objectives

  • Placements/formats: How platforms deliver ads across different ad placements and formats

Ultimately, the ROI generated by these delivery levers is what matters most to advertisers, but selection and survivorship biases make comparing ROI across platforms difficult. We believe marketshare is the best proxy for each platform’s advertiser ROI.

1. Google: For all of the above delivery levers, Google leads the industry. Consistently, Google provides simplified delivery options for less sophisticated advertisers while providing robust, flexible delivery options for sophisticated and enterprise customers.

For example, Google offers eight different bidding strategies (including strategies that Facebook doesn’t offer, such as bid optimization for ad position or portfolio-level bidding) but automates bidding by default for most advertisers. While Google offers budgeting solutions similar to Facebook (e.g. automatic budget allocation and accelerated delivery), Google defaults to providing advertisers with budget options rather than open-ended budgets – helping advertisers understand the implications across a range of budgets.

For audience targeting, Google offers advertisers unique targeting options such as content-specific and purchase intent-specific targeting. Google is also more ambiguous about whether they utilize targeting as a signal or a constraint. Though Google’s placement and format options are more complex than Facebook’s, Google has created verticalized and simplified solutions that do not require an advertiser to make specific placement decisions to execute their campaign goals.

2. Facebook: As the obvious but distant second place in delivery lever sophistication, Facebook offers most but not all of Google’s delivery options but fails to tailor the experience for advertiser segments as successfully as Google does. For example, Facebook’s flexible but complicated bidding and budget options are hidden from LWI users but are not comparatively simplified for advertisers with a medium level of sophistication.

To many advertisers, Facebook is perceived as the leader in audience targeting (due to the perception that Facebook knows most about customers and enables detailed targeting options to make use of the customer data). Still, this perception often leads to less sophisticated advertisers overusing targeting features. Similarly, while Facebook simplifies advertising across placements and offers an array of visually appealing formats, placement optimization is comparatively tedious. Facebook does not have an “LWI” for less sophisticated advertisers who want to advertise across both FB and IG.

3. Microsoft: Microsoft offers delivery levers with similar options as Facebook but rarely simplifies or hides those options for a given advertiser segment. For example, Microsoft and LinkedIn offer similar bidding capabilities as Facebook, and Microsoft even offers a campaign budget optimization-like feature called ‘shared’ budgets. The platforms provide targeting capabilities such as remarketing, ‘similar audiences’ (Microsoft’s lookalike audience offering), custom audiences, LinkedIn profile targeting, and product audiences. Still, the products lack (or are perceived to lack) Facebook’s interest targeting granularity. While Microsoft and LinkedIn have linked some targeting features, they have not unified their ad buying platforms, making ad placement/format management across properties more difficult.

4. Twitter: Like Microsoft, Twitter offers similar delivery lever options as Facebook but (aside from within their promoted tweets feature) largely fails to tailor the experience to advertiser segments. While Twitter offers similar bidding and budget capabilities as Facebook, Twitter differentiates itself by embracing cost per outcome-bidding – enabling advertisers to pay per view, install, click, engagement, or follow. Similarly unique are Twitter’s targeting features, which focus around Twitter-specific use cases (e.g. the ability to target real-time keywords, topics, events, and TV shows). Twitter also offers Twitter Promote Mode to enable advertisers to extend the reach of their tweets for a flat monthly rate. Still, due to the nature of their platform, Twitter has limited placement/format options.

5. TikTok: TikTok’s delivery levers closely resemble Facebook’s in both UX and content. But, likely due to the platform’s relative immaturity, the platform lacks many of Facebook’s more advanced features. For example, TikTok offers optimized cost per click bidding but lacks cost cap bidding and a wider array of objectives. TikTok offers scheduled and dayparted budgets but lacks campaign budget optimization. While TikTok has granular interest targeting, user interests are inferred from fewer user touch points (though the app does offer a unique ‘broad targeting’ option that uses interests as only inputs rather than constraints).

6. Amazon: Amazon has the fewest delivery lever options but – since their ads are primarily sponsored product ads for their own marketplace – the platform requires fewer delivery lever options. For bidding, Amazon offers an unintuitive ‘dynamic bids’ feature wherein the system will either increase/decrease or just decrease your bid based on whether a conversion is likely. For budgets, Amazon offers only a daily budget but does provide portfolio budget limits. Amazon’s targeting options are sparse – with a primary focus on keyword and product targeting.

Competitive audit: Automation

Ad platforms exhibit varying levels of automation, too. Automation simplifies advertiser workflows, helps advertisers discover new tools and solutions, and eliminates low performance outliers. Still, automation can be a double-edged sword – sophisticated advertisers often prefer more control over more automation, and some implementations of automation can decrease advertiser trust (e.g. fear of the unknown). We analyzed each competitor’s automation solutions for:

  • Delivery lever automation: How each platform automates (or provides the option to automate) decisions for bidding, budget management, audience targeting, campaign optimization, and placements/formats

  • Creative automation: How each platform automates creative generation – particularly as it relates to real-time ad personalization

1. Google: Google not only automates the most across delivery levers but also provides sophisticated advertisers with manual controls when desired (except for app advertisers who all must use Google’s automated solution). In fact, for Smart Campaigns (Google’s optional experience for SMBs) and App Campaigns (Google’s mandatory experience for app advertisers), Google is the only platform to automate campaign structure decisions. Campaign configuration, targeting, placements, and bidding options have been completely removed from these products.

For example, with Smart Campaigns, businesses simply state their goal (restricted to calls, store visits, website actions), ad copy, budget, and location. Google then automates and optimizes the ad. Elsewhere, Google's Recommendations tool will even automate campaign management/clean-up via the click of a button. Google also offers automated verticalized solutions that align with Google’s Cloud offering.

On the creative automation front, Google offers the most advanced creative tools and is most aggressive with enabling creative automation by default (such as dynamic creative as the only option for Smart Campaigns). In addition to offering its own versions of what all other platforms have, Google also has unique creative automation tools such as its features to simplify asset generation (targeted at SMBs), auto-upload creative assets from an advertiser’s website, and customize headlines on the fly.

2. Facebook: Facebook lacks many of Google’s automation features, including campaign structure automation and automated recommendation resolution. While FB recently released a solution for automated app ads, Google provides this level of automation for all advertisers (if they choose to use it). Still, Facebook provides comparable optimization options for all other delivery levers (but does not package them as a single solution) and stands out among other competitors with regard to creative optimization (e.g. dynamic ads and dynamic creative).

3. Microsoft: Microsoft and LinkedIn lag Facebook’s automation due to lack of targeting, placement, and budget allocation automation. Microsoft does offer a complex suite of automated rules to change keyword bids automatically, monitor an ad group’s performance, and start and stop your campaign at preset times. From the perspective of liquidity, LinkedIn’s automated placements as the only option is quite different from Facebook’s more flexible approach. While Microsoft offers dynamic creative features for search ads similar to Google, it has no display ad automation features. Meanwhile, LinkedIn offers very rudimentary display automation relative to competition: it does not customize creative based on what the user is most likely to convert from but instead based on simple information from the user’s profile.

4. Twitter: Twitter lags FB with regard to lack of placement automation and budget allocation automation. Twitter does offer autobid and campaign optimization bidding strategies, and for targeting, Twitter offers conversation targeting (where advertisers can easily target audiences based on the conversations they're actively participating in on Twitter). Twitter also has lookalike audiences. For smaller advertisers, Twitter offers 'Promote Mode', which automatically boosts posts for a fixed fee per month. Otherwise, Twitter does not offer creative automation tools but does rotate/select the best creative automatically.

5. TikTok: TikTok lacks much of the automation that advertisers on Facebook rely on, such as automated bidding, targeting, and budget allocation. While TikTok does not offer automated targeting, they have effectively treated the default targeting as automated through use of oCPC. TikTok does have automatic placements across TikTok, TikTok audience network, and News Feed apps. TikTok greatly lags the competition for creative automation.

6. Amazon: Amazon has no differentiating automation features and is the only major player without automated bidding. Amazon does not offer creative automation tools, but they recently acquired Sizmek, who offers rules based creative optimization (e.g. location-specific creative, pricing, products). Still, due to the relatively few decisions Amazon advertisers must make, Amazon requires less automation. And since most ads are sponsored products, creative automation is not necessarily required either. Due to peculiarities in how the platform handles ad groups, the platform would be well-suited to implement campaign structure or budget automation.

Competitive audit: Performance guidance

Ad platforms often provide performance guidance to help advertisers make decisions that optimize performance. The performance guidance on these platforms either supplement or replace automated features and exhibit varying levels of personalization, assertiveness, and effectiveness. This analysis reviewed three types of performance guidance:

  • Performance recommendations: How ad platforms advertisers make the decisions that optimize their performance

  • Performance diagnostics: How ad platforms help advertisers diagnose performance issues

  • Expectation-setting tools: How ad platforms help advertisers predict and set expectations about future performance

1. Google: Google provides the most prescriptive, comprehensive, and pervasive guidance in the industry. In particular, Google’s recommendations platform stands out amongst competitors. In addition to having the largest recommendation inventory (at least 70 unique recommendations), Google has user-friendly recommendation filtering and dismissal, recommendation estimates (7 day predictions for CPA, results, and cost), and even an optimization score to gamify recommendation resolution. Advertisers can also resolve recommendations in bulk.

Google’s expectation-setting tools are similarly sophisticated. Google’s Keyword Planner is (alongside Microsoft’s clone) the only planning tool that enables long-term estimates with custom date ranges. Uniquely, this tool even lets advertiser input components of the estimate calculation (e.g. conversion rate). Unlike competitors, Google’s bid and budget simulators appear in context (e.g. next to a Limited by Bid ad), so they are easily available when they are needed. While most ad platforms provide daily estimates, Google’s performance estimates are weekly (which is preferred by advertisers). Even Google’s automated solutions offer expectation-setting tools for the few decisions where advertisers must make decisions (e.g. budget).

Google seamlessly integrates its recommendations and expectation-setting tools with its wide breadth of performance diagnostic tools. The diagnostics typically indicate the best path forward and are often linked with Google’s Bid Simulator. Diagnostics are gamified, too – Google even offers an ‘ad strength’ score at the ad level to spur adoption of best practices.

Google’s expectation-setting tools are similarly sophisticated. Google’s Keyword Planner is (alongside Microsoft’s clone) the only planning tool that enables long-term estimates with custom date ranges. Uniquely, this tool even lets advertiser input components of the estimate calculation (e.g. conversion rate). Unlike competitors, Google’s bid and budget simulators appear in context (e.g. next to a Limited by Bid ad), so they are easily available when they are needed. While most ad platforms provide daily estimates, Google’s performance estimates are weekly (which is preferred by advertisers). Even Google’s automated solutions offer expectation-setting tools for the few decisions where advertisers must make decisions (e.g. budget).

Google seamlessly integrates its recommendations and expectation-setting tools with its wide breadth of performance diagnostic tools. The diagnostics typically indicate the best path forward and are often linked with Google’s Bid Simulator. Diagnostics are gamified, too – Google even offers an ‘ad strength’ score at the ad level to spur adoption of best practices.

Still, Google is not without its performance guidance weaknesses. We’ve heard from advertisers that say that Google’s recommendations are primarily upsells that are “in the best interest of Google, not me.” Advertisers lose trust in Google’s guidance over time (especially those for budget) as advertisers begin to perceive that the guidance does not align with each advertiser’s true goals.

2. Microsoft: Microsoft’s performance guidance solutions are largely clones of Googles (which are admittedly the best in the industry). Like Google, Microsoft offers an extensive recommendation suite complete with estimates, dismissal features, and categorization. Microsoft provides long-term expectation-setting tools for keyword/search advertising but lags competitors on the display ad front.

LinkedIn also offers unique expectation-setting tools. Their version of Facebook’s Estimated Daily Results (which most competitors have mimicked) also has 7-day and 30-day estimates, and it even allows advertisers to break down estimates by audience type. LinkedIn also estimates click-through rate.

3. Facebook: Facebook still lags the performance guidance features that Google launched in 2018 (and that Microsoft has since cloned). While Facebook now has performance recommendations, the inventory is small (just five recommendations), the feature is available for just 50% of sophisticated advertisers, the recommendations lack estimates and bulk resolution flows, and there are no dismissal or categorization features. Unlike Google, the recommendations that Facebook does have are not as prescriptive. For example, while Facebook will recommend to increase budget, Google will recommend the new budget to use.

While Facebook provides estimated daily results, Google provides 7-day results in context (e.g. when performance is limited and the advertiser is considering a change). Facebook also lacks longer-term prediction tools for auction buying (though they are available for R&F).

Facebook’s performance diagnostics tools (e.g. ad relevance diagnostics, Learning Limited) are on par with Google’s, but Google has developed gamified diagnostics to extend the scope of performance guidance beyond when changes obviously improve CPA. Unlike Google, Facebook also lacks suggested resolution flows for a given diagnostic tool (e.g. how do I improve a low score?).

Competitors such as LinkedIn, TikTok, and Twitter also cite Facebook’s lack of audience breakdown tools – a feature often used for performance diagnostics. On these platforms, it is possible to break down past performance by audience interest. While Facebook avoids such features to reduce the breakdown effect, advertisers nonetheless describe their Facebook ad strategy as more fragmented (with higher ad volume) than competitors, because advertisers get around our breakdown limitations by creating more ad sets.

4: Twitter: Aside from limited guidance during ad creation, Twitter provides very little performance guidance for advertisers. First, other than suggested bids, Twitter does not provide recommendations. Second, other than Twitter’s clone of Estimated Daily Results and Potential Reach, Twitter does not offer expectation-setting tools.

Still, Twitter’s breakdowns by interest feature is not possible on Facebook, and this contributes to advertisers considering Facebook a “black box”. Twitter explicitly makes this comparison in Sales materials.

5. TikTok: TikTok’s performance guidance offerings closely resemble Twitter’s. TikTok completely lacks recommendations. Aside from a similar clone of Facebook’s Estimated Daily Results and Potential Reach, TikTok also lacks expectation-setting tools. On the other hand, TikTok has mimicked Facebook’s learning phase diagnostic (though it lacks the Learning Limited concept) and provides performance breakdowns by interests similar to Twitter.

6. Amazon: As a simpler platform with a more homogeneous advertiser base, Amazon requires less performance guidance. Amazon offers suggested keywords and bids but otherwise provides no recommendations. The suggested bid column shows a recommendation for how much you should bid on a keyword and has an ‘Apply’ button in-line. Otherwise, Amazon has few performance guidance features, but the primary metric for DR — ACoS — is incredibly important for their clientele and more accurate here than when reported on 3rd party platforms.

Conclusion and next steps

In assessing Google’s dominance, we identified opportunities for Facebook within delivery levers, automation, and performance guidance. Facebook should consider further tailoring its delivery lever options for each advertiser’s sophistication level, uniting automated solutions as a one-stop shop, and rapidly accelerating performance recommendation investments. Even Amazon (who demonstrates how simple a sponsored product ad platform can be) and Twitter (who embraces audience interest breakdowns in contrast with Facebook) suggest future opportunities for Facebook.

To further enhance our knowledge of competitors across relevant dimensions for the delivery organization, we will also be exploring their capabilities within intent and value exchange. As much of the work around the advertiser intent taxonomy and segmentation has been recently published, we did not have existing competitive inbound to include in this analysis. However, based on our early observations, this is a space where Facebook has an opportunity to clearly differentiate and deliver more value for advertisers. For example, all platforms only enable goal expression at the campaign level and do not differentiate between a true business goal and a marketing goal. In addition, restrictions must be expressed through standard delivery levers such as geographic or demographic targeting.

Finally, many questions emerged from this initial audit that may further help us to identify and focus our efforts on product initiatives that will deliver the most value. These include:

  • Forward looking analysis: Now that we have an understanding of the current state of these platforms, where do we think each platform is headed in the future?

  • Opportunity prioritization: Of the various opportunities identified, which ones might advertisers feel would deliver the most value?

  • Segment-driven audit: How are our competitors delivering experiences for different segments of advertisers?

We’ll consider these questions as we utilize these insights to inform product prioritization and identify new inbound needs.

8 views0 comments

All work © Jocelyn Jiang 2014 - 2021 • all rights reserved