Search Advertising
Structure, relevance, and bidding
Search advertising buys visibility at the moment of intent. Account structure should mirror business priorities: campaigns for budget and targeting control, ad groups for tightly themed keyword sets, and ads that echo the query. Understand match types - exact, phrase, and broad - and pair broad reach with disciplined negative keyword lists and search term reviews. Quality Score summarizes expected CTR, ad relevance, and landing page experience; Ad Rank combines bid with quality and expected impact of extensions, so a lower bid can outrank a higher one.
Choose bidding strategy against data volume: manual or maximize-clicks early, then target CPA or ROAS once conversions are stable. Responsive search ads need distinct headlines and descriptions covering benefits, proof, and calls to action; extensions add real estate and relevance. Landing page congruence is a ranking and conversion factor. Conversion tracking must be accurate and deduplicated, and attribution windows should be understood before drawing conclusions. Performance Max and broad match can expand reach but require strong signals and exclusions.
Traps candidates fall for
- Launching broad match without negatives or search term hygiene.
- Equating Quality Score with Ad Rank.
- Assuming average CPC equals the marginal cost of the next click.
- Switching to target CPA before enough conversions exist.
- Ignoring assisted conversions and view-through effects.
Worked example
Two advertisers bid $2.00 and $3.00 with quality scores of 10 and 4. Using bid × quality as a simplified Ad Rank proxy, the first scores 20 and the second 12, so the lower bidder wins. Rule of thumb: improve relevance and landing page experience before raising bids. For profitability, compare ROAS with margin: $10,000 revenue on $2,000 spend is 5:1 ROAS, but at 30% margin, gross profit is $3,000 against $2,000 spend.
Email Marketing, Automation, and Conversion Optimization
Consent, relevance, and lifecycle design
Email performance rests on permission and relevance. Build lists through explicit opt-in, honor preferences, and maintain hygiene with double opt-in where appropriate. Deliverability depends on authentication - SPF, DKIM, and DMARC - plus consistent sending, low complaint rates, and clean list management. Segment by behavior, lifecycle stage, and value rather than blasting one message to everyone. Lifecycle automation covers welcome, nurture, abandonment, re-engagement, and post-purchase sequences; lead scoring decides when a contact moves to sales.
Conversion rate optimization is disciplined experimentation. Map the funnel, identify friction through analytics, session replay, and user testing, then prioritize by expected impact and effort. A/B tests need one isolated variable, a pre-registered primary metric, sufficient sample size, and a decision rule. Multivariate testing suits high-traffic pages with several independent elements.
Traps candidates fall for
- Treating open rate as reliable; privacy features inflate it through image proxies.
- Batch-and-blast sending with no segmentation or suppression.
- Calling a test a winner before reaching statistical confidence.
- Changing several variables at once and attributing the result to one.
- Optimizing a step in isolation while shifting friction downstream.
- Using dark patterns that lift short-term conversion and damage trust.
Worked rule of thumb
Before testing, state hypothesis, metric, minimum detectable effect, and sample size. Example: a checkout page with 20,000 monthly visitors tests a simplified form. Baseline conversion is 3.0%; the test reaches 3.4% with adequate sample and 95% confidence, so the team ships it only if revenue per visitor rises after monitoring refunds and support contacts. For email, measure revenue per recipient and unsubscribe rate alongside clicks, not open rate alone.