The ultimate guide to performance marketing | Timothy Davis (Shopify)
Timothy Davis, former Head of Performance Marketing at Shopify, shares tactical insights on paid growth. He covers when to invest, how to run "signs of life" tests, platform choices, common agency mistakes, and building an internal team.
Deep Dive Analysis
18 Topic Outline
Defining Performance Marketing and Its Scope
Why Paid Search is Essential for Every Business
Identifying Growth Potential and Emerging Channels
How to Experiment with New Ad Platforms (Signs of Life Tests)
Choosing Core Ad Platforms: Google, Meta, YouTube, LinkedIn
When to Start Investing in Paid Marketing (Pre-Product Market Fit)
Common Mistakes Agencies Make and How to Avoid Them
Hiring Your First Performance Marketing Specialist
Key Metrics and Reports for Google Ads Performance
Understanding Impression Share vs. Click Share
Identifying True Competitors with Advanced Metrics
Multi-Touch Attribution and Incrementality Testing
Building and Scaling an Internal Performance Marketing Team
Effective Training and Onboarding for New Hires
Impact of Apple's ATT on Paid Marketing
The Underestimated Power of Ad Creative
AI's Current and Future Role in Performance Marketing
Forward Thinking, Backward Planning for Growth
9 Key Concepts
Performance Marketing
Performance marketing is any marketing effort where the performance can be measured, typically online. It focuses on quantifiable results, though it can also encompass offline efforts and affiliate marketing.
Paid Search
Paid search is a user-driven form of advertising where ads appear in response to specific keywords typed by a user. It's considered a baseline for almost every business because it targets users with existing intent.
Disruptive Media
Disruptive media refers to advertising that interrupts a user's current activity, such as seeing an ad on social media while browsing or during a video. Unlike paid search, it's not initiated by the user's direct search query.
Signs of Life Test
A 'signs of life test' is a small, low-budget experiment on a new ad platform to determine if there's any positive signal or potential for a business. If initial positive results are observed, a more comprehensive campaign can then be developed.
Lookalike Audiences
Lookalike audiences are created by uploading an existing customer base to an ad platform, which then identifies other users with similar behaviors or characteristics. The percentage (e.g., 1% to 10%) indicates how closely these new users resemble the original customer base.
Signal vs. Noise
In the context of data, 'signal versus noise' refers to the ability to identify truly important and actionable data points ('signal') from the overwhelming amount of less relevant information ('noise') provided by ad platforms. This distinction is crucial for effective optimization.
Multi-Touch Attribution
Multi-touch attribution is a method of assigning credit to all marketing channels a user interacted with before converting, rather than just the first or last touchpoint. A common model is time decay, which gives more weight to interactions closer to the conversion event.
Incrementality
Incrementality measures the additional growth or conversions a marketing campaign generates that would not have occurred without that specific ad. It helps determine the true value added by paid advertising beyond what would have happened organically.
Ops Cadence
An 'ops cadence' is a structured, scheduled routine, often documented in a spreadsheet, that outlines the frequency and type of performance marketing activities (e.g., keyword updates, ad copy reviews, search query reports) performed by a team. It ensures accountability and consistent execution.
14 Questions Answered
Yes, paid marketing is for everyone, especially paid search, because platforms like Google and Meta increasingly prioritize paid listings, making organic reach more challenging.
Companies should analyze their existing data to see where users are already finding them (e.g., TikTok, Google Analytics) and then 'turn that knob up to 11' with paid ads on those successful channels.
Start with a 'signs of life' test by using your existing customer data to build highly targeted lookalike audiences (e.g., 1% match on Meta) and run a small, limited budget campaign to gauge initial positive signals.
Start by checking if the target audience is correct and engaging (click-through rate). If engagement is low, it could be the creative. For deeper insights, focus groups or A/B tests with different creative elements can help understand user resonance.
Generally, start with Google Search due to its user-driven nature, then move to Meta (Facebook and Instagram). If video creative is available, YouTube is highly recommended. LinkedIn is powerful but more expensive, suited for B2B or high LTV products.
No, it's generally not recommended to run significant paid ads before product-market fit, especially if operational issues (like currency support) prevent conversions. Doing so can annoy users and lead to poor conversion rates and negative brand experiences.
Agencies often use generic playbooks, don't delve deep into account specifics, and are too busy with many clients to focus on granular optimizations like landing page experience, conversion funnels, or specific ad relevance improvements.
Agencies are good for starting out, but once monthly ad spend reaches a certain threshold (e.g., $50K/month), it's time to start conversations about hiring an internal specialist who can dedicate full attention to the company's specific needs and data.
The most crucial quality is a strong data-driven thought process and the ability to distinguish 'signal' from 'noise' in the vast amount of data provided by ad platforms. Technical skills for running ads can be taught.
Leverage platform partners (Google, Meta, LinkedIn) who can provide anonymized industry benchmarks for metrics like CPC and click-through rate. Additionally, use tools like Google's auction insights to identify 'true competition metrics' and genuine threats.
Adopt a multi-touch attribution model, with a preference for time decay, which gives more credit to recent interactions. Be aware that attribution models are inherently biased and don't account for incrementality.
Conduct geo-experiments (geo-ex) or conversion lift tests directly within the ad platforms. These experiments compare a control group (not shown ads) to a test group to determine the actual uplift in conversions that wouldn't have occurred otherwise.
As long as platforms can use SKAdNetwork (SKAN) to provide attribution and measurement for iOS campaigns, performance marketers can continue to operate effectively, adapting to the changes in tracking.
Ad creative is hugely impactful and often underestimated. Great creative tells a story and evokes emotion (comedy, happiness), leading to a lasting impression and more favorable user action.
29 Actionable Insights
1. Embrace Paid Marketing
Consider paid marketing for every company, especially paid search, as platforms increasingly prioritize paid listings over organic results. Paid search is user-driven, making it a baseline investment for nearly all businesses.
2. Focus on Data Signal
When hiring for performance marketing, prioritize candidates who can distinguish ‘signal’ from ’noise’ in data, as data analysis is more critical and harder to teach than platform mechanics. This ensures focus on relevant metrics that drive business results.
3. Forward Think, Backwards Plan
Define your ultimate long-term goals (forward thinking) and then meticulously map out the immediate, iterative steps and milestones required to achieve them (backwards planning). This strategic approach ensures progress towards ambitious objectives.
4. Test for Signs of Life
Conduct small, low-budget ‘signs of life’ tests on new platforms to identify early positive signals before scaling. If a signal is found, then build a comprehensive campaign around it, rather than immediately going full throttle.
5. Prioritize Product-Market Fit
Avoid significant paid ad investment if your product lacks market fit or operational readiness (e.g., correct currency support). Running ads without this foundation will likely lead to low conversions, wasted spend, and user annoyance.
6. Align Metrics to Goal
Focus on specific metrics that directly align with your campaign’s primary objective, such as conversions for purchase, reach for awareness, or funnel entry for consideration. Disregard metrics that are irrelevant to the campaign’s core goal to avoid noise.
7. Build Lookalike Audiences
Initiate platform experiments by uploading your existing customer base to build 1% lookalike audiences. These audiences are highly correlated to your target users and offer the strongest initial signal for potential success.
8. Start with Google Search
Begin your paid growth efforts with Google Search due to its user-driven nature, where users actively search for relevant keywords. Once established, expand to Meta (Facebook/Instagram) and YouTube, especially if you have video creative available.
9. Leverage Platform Partners
Consult platform representatives (Google, Meta, LinkedIn) for anonymized competitive benchmarks and industry insights. They can provide valuable data on metrics like click-through rate and CPC for your specific competitors, helping you understand your performance relative to the market.
10. Cultivate Learning Mindset
Create an organizational culture where it’s acceptable to fail in experiments, viewing every outcome as either a win or a learning opportunity. This reduces pressure and encourages continuous experimentation and improvement.
11. Optimize Ad Relevance
Continuously improve ad relevance and strength by incorporating platform suggestions, such as adding unique headlines or keywords. This directly boosts your quality score, leading to more impressions and better performance.
12. Eliminate Poor Performing Ads
Proactively identify and pause or remove ads with poor quality scores or high CPCs, as even a single underperforming ad can negatively impact overall account performance. This improves the account-level quality score and efficiency.
13. Focus on Click Share
Prioritize showing ads to the right person as often as possible, measured by click share, rather than merely maximizing overall impressions. This shifts focus from ’ego marketing’ to effective user engagement and conversion.
14. Strategic Competitor Bidding
Consider ‘coattail riding’ by bidding on competitor keywords if your product offers a similar solution. However, ensure your ad copy adheres to trademark rules to avoid violations and potential removal by platforms.
15. Use Multi-Touch Attribution
Employ a multi-touch attribution model, preferably time-decay, to credit all channels that contribute to a user’s conversion journey. This acknowledges that initial touchpoints are less impactful for conversion than later interactions.
16. Test Creative Iterations
Always test different ad copy and creative variations, even seemingly simple ones, as the most effective options may not be what you initially expect. No idea should be left on the cutting room floor; always be willing to learn what works.
17. Invest in In-House Creative
Consider hiring an in-house creative person for better brand tone matching, faster iteration, and consistent content production. In-house talent often outperforms agencies in matching brand identity and quickly generating new test ideas.
18. Implement Ops Cadence
Establish a detailed operational spreadsheet (ops cadence) outlining weekly, bi-weekly, and monthly tasks for all performance marketing activities. This ensures accountability, provides transparency to cross-functional teams, and helps prevent forgotten tasks.
19. Accelerate New Hire Impact
Onboard new team members quickly by providing clear expectations, early responsibilities, and hands-on training, aiming for impact within 30-45 days. This approach gets people up to speed faster than slow-rolling introductions and handbook reading.
20. Evaluate Hiring Needs
Use a workload calculator to track team members’ time spent on meetings, PTO, optimizations, and reporting. Hire new team members only when existing staff consistently exceed capacity after efforts to cut non-essential work and streamline processes.
21. Test Creative with Control
When testing new creative, run A/B tests with a control group (e.g., a logo-only ad or existing ad) against your new creative. This helps isolate the creative’s true impact on engagement and conversion, providing clearer insights.
22. Leverage Video Ads
Prioritize video advertising on platforms like YouTube if you have the creative resources, as video is currently performing very well when measured correctly. Ensure you have a ‘flywheel’ for consistent creative refresh.
23. Craft Emotional Creative
Develop video ads that evoke emotion, such as comedy or happiness, to create a memorable connection with users. Emotional ads are more impactful and lead to more favorable actions after viewing.
24. Consider LinkedIn for B2B
Use LinkedIn for B2B marketing, especially for higher LTV products, due to its powerful targeting capabilities (job titles, companies). However, it is more expensive, so establish a presence on Google and Meta first.
25. Defend Against Brand Bidding
Monitor for competitors bidding on your trademarked brand terms; if found, report violations to Google and analyze auction insights for genuine threats versus accidental close variants. Google will disallow trademarked names in ad copy, but misspellings can slip through.
26. Utilize In-Platform Tools
Maximize the use of native platform tools and features, as they directly influence the signals sent to the platform’s algorithms for optimization. Sending the right signal helps the platform deliver desired business results.
27. Run Incrementality Tests
For significant ad spend (over $50K/month), partner with platforms to run geo-experiments or conversion lift tests to measure the true incremental impact of campaigns. This ensures you’re not taking credit for conversions that would have happened anyway.
28. Collaborate on Ad Copy
Foster collaboration between performance marketers and creative teams for ad copy, testing all ideas. The performance marketer brings data insights, while the creative team offers fresh perspectives, leading to more effective ads.
29. Hands-On Account Management
Ensure performance marketers are actively managing accounts, making weekly, bi-weekly, and monthly adjustments directly within the ad platforms. This ‘getting shit done’ approach drives the best results by staying on top of optimizations.
7 Key Quotes
Hot take, paid is for everyone.
Timothy Davis
I believe in creating an environment where it's okay to fail because we're either winning or we're learning.
Timothy Davis
I'm a firm believer in hiring smart people and then getting out of their way.
Timothy Davis
Who cares about how many impressions you're getting? Who cares about the reach? Who cares about the frequency? If the goal of the campaign is to get the conversions, that's what we should be doing.
Timothy Davis
The biggest thing I want to focus on is what is your thought process when it comes to data? Because I can teach anyone how to do Google ads. I can teach anyone how to do meta ads. That is not the hard part. It is the data part that is the hard part.
Timothy Davis
Happiness is dedicated by expectations or dictated by expectations.
Timothy Davis
You won't see it for what it is until you stop looking through the lens of what you want it to be.
Timothy Davis
5 Protocols
Experimenting with a New Ad Platform (Signs of Life Test)
Timothy Davis- Use your existing customer base data to build lookalike audiences within the platform.
- Start with a highly correlated lookalike audience (e.g., 1% match on Meta) for your initial test.
- Allocate a very small, limited budget to this test.
- Observe if there's a positive signal or 'sign of life' from the initial results.
- If a positive signal is found, then build a comprehensive campaign with appropriate creative and messaging, rather than immediately scaling up.
- Give yourself grace; it's okay if it doesn't work, as long as you learn from the experiment.
Building a Performance Marketing Team (First 3 Hires)
Timothy Davis- First Hire: A data-driven individual who can find 'signal in the noise' within ad platforms and understands data thought processes. This person will wear multiple hats (e.g., Growth Marketing Specialist/Manager).
- Second Hire: A creative person, specializing in graphic design and branding, who can produce ads matching the brand's tone and performance goals, iterating quickly on new ideas.
- Third Hire: A dedicated data scientist to assist with incrementality testing, creating advanced reports, and building analyses that generalists cannot.
Managing Team Workload and Hiring Strategy
Timothy Davis- Use a calculator (spreadsheet) to track how much time each team member spends on meetings, PTO, optimizations, reporting, and other tasks.
- Estimate the total number of days required for committed tasks versus actual days available in the quarter.
- If a team member is consistently 'in the red' (more work days than available days) for multiple quarters, investigate.
- First, identify if tasks can be cut or if team members can reduce meeting attendance or involvement in certain launches.
- If, after cutting tasks, the team member is still in the red, then begin the process of hiring a new resource, clearly defining their responsibilities and expected workload.
Effective Training and Onboarding for New Hires
Timothy Davis- Provide clear expectations and responsibilities early on, aiming for impact within 30-45 days (instead of the typical 90).
- Share an 'ops cadence' (spreadsheet) that outlines how often specific performance marketing activities are done within the team.
- Offer hands-on training by opening accounts and demonstrating efficient workflows, even remotely.
- Give new hires ownership of a campaign or project early to foster responsibility and quicker learning.
- Do not expect perfection immediately; be open to questions and learning from new hires.
Forward Thinking, Backward Planning for Growth
Timothy Davis- Define your ultimate growth goals (e.g., performing on all platforms, achieving 'always-on' status for an emerging channel).
- Determine what it would take for a channel to be considered 'performant' (e.g., a specific number of conversions per month, spend, and lift).
- Backward plan from that goal, identifying the micro-milestones and iterations needed to get there.
- Start with foundational activities (e.g., paid search content and keywords) and gradually add more complex elements (e.g., creative-driven channels).
- Continuously iterate and adapt your strategy based on learnings from each milestone.