Digital Marketing and Online Advertising

01Purpose and scope

• This manual supports the course Digital Marketing and Online Advertising. It describes how practitioners plan, launch, monitor, and refine campaigns that reach people through search results, social feeds, video platforms, email, websites, and partner sites. It is written as a working reference for marketers, analysts, account coordinators, and small business owners who run their own promotion.
• The scope covers paid media, the owned channels that support it, and the measurement that connects spending to outcomes. It does not cover detailed graphic design, software engineering for tracking code, or legal review of claims and privacy notices. Those areas have their own specialists and their own standards, and a marketer should know when to hand work to them.
• Practice varies a great deal by context. A regional bakery with a fixed delivery radius and a national insurer with heavily regulated product claims face different constraints, different approval paths, and different tolerance for risk. This manual describes common practice and points out where those differences matter.

02Core principles

• Start from a business outcome, not a channel. A campaign exists to produce a specific result, such as qualified leads, trial sign-ups, store visits, or repeat orders. Platform metrics such as clicks or impressions are useful signals, but they are not the goal. When a channel looks busy but does not move the outcome, the channel is not working for that goal.
• Define the audience by behavior and context when possible. Demographic targeting is available on most ad platforms, yet intent signals such as a recent search, a visit to a pricing page, or membership in a relevant interest group often predict response better. Use the narrowest audience that still reaches enough people to learn from.
• Treat every campaign as a test with a stated hypothesis. Write down what you expect to happen, why, and what result would change your mind. This habit prevents post hoc explanations, where every outcome is described as a success.
• Keep the customer experience consistent from ad to landing page to follow-up message. A discount promised in an ad should appear on the page without friction. Mismatched promises raise cost per acquisition and damage trust, even when the ad itself performs well on clicks.
• Respect the limits of attribution. No model tells you exactly which touchpoint caused a sale. Last-click, first-click, and data-driven models each assign credit differently and each has blind spots. Use more than one lens, and state the lens you used whenever you report results.

03Planning and preparation

• Begin with a brief that names the product or offer, the target action, the audience segments, the geography, the budget range, the launch window, and the approval owner. A brief that cannot be written in one page usually signals an unclear objective. Send the brief to stakeholders before any account changes are made, and keep the version history.
• Audit the existing setup before adding new spend. Check conversion tracking, tag firing, audience lists, negative keyword lists, and any active campaigns that overlap with the new one. Overlapping campaigns often bid against each other and inflate costs. A common first-week task is confirming that the conversion event recorded in the ad account matches the event the business actually values.
• Choose channels based on where the audience spends time and how the offer is shaped. Search ads capture people who already want something specific, so they suit high-intent products such as plumbing services or software trials. Social and video placements are better for creating interest in a new product category, though they usually produce lower immediate conversion rates. Email and retargeting work well for people who already know the brand.
• Set the budget with a clear view of trade-offs. Daily budgets, lifetime budgets, and bid strategies each change how the platform spends money. Automated bidding can improve efficiency once enough conversion data exists, but it can also drift if the data is thin or the conversion event is weak. Reserve part of the budget for testing, and state what share that represents.
• Prepare creative and landing pages in parallel with the account setup. Ad copy often needs several variants so that the platform can learn which messages perform. Landing pages should load quickly on mobile devices, state the offer above the fold, and include a single primary action.

04Recommended workflow

1. 1
   : Confirm the objective, the primary metric, and the guardrail metrics in writing. The primary metric might be cost per qualified lead, while guardrails might include maximum cost per acquisition and minimum lead quality score from the sales team.
2. 2
   : Build the measurement layer before launch. Verify that conversion events fire once per real action, that duplicate events are filtered, and that the test environment shows the expected values. Run a short test with a small budget and check that the numbers in the ad account match the numbers in the web analytics tool within a reasonable tolerance.
3. 3
   : Structure campaigns by intent and audience rather than by convenience. Group keywords or audiences that share a message and a landing page. Avoid putting unrelated products in one ad group, because the platform will show one message to people who need different ones.
4. 4
   : Write ad variants that test one meaningful difference at a time. Change the headline promise, the call to action, or the proof point, but not all three at once, unless you are deliberately testing a complete concept. Label each variant so the results are readable later.
5. 5
   : Launch with conservative bids and a daily budget that allows the platform to exit its learning phase. Avoid editing bids, budgets, or targeting every few hours, because frequent changes reset learning in many automated systems. Make a schedule for reviews, such as every two or three days in the first two weeks.
6. 6
   : Review search terms, placements, and audience performance at regular intervals. Add irrelevant search terms as negatives, exclude placements that generate clicks without conversions, and shift budget toward segments that meet the guardrails. Record each change with the reason and the date it was made.
7. 7
   : Report results against the stated hypothesis. Present the primary metric, the guardrail metrics, the attribution lens used, and the sample size. State whether the test reached a conclusion or whether more data is needed, and recommend the next action clearly.

05Quality checks, risk controls and compliance

• Check every ad against platform policies before submission. Ad platforms review copy, images, landing pages, and targeting for restricted categories such as financial products, health claims, alcohol, gambling, and political content. Policies change frequently, so consult the platform's current policy center rather than relying on memory or a colleague's notes. Rejected ads waste time, and repeated violations can restrict an account.
• Review privacy obligations before using personal data for targeting or retargeting. Consent rules, data retention limits, and opt-out requirements vary by country, by state, and by the type of data involved. Some jurisdictions impose strict requirements on tracking cookies and on uploading customer lists to ad platforms. Confirm current requirements with your legal or privacy team, and treat the relevant authoritative regulator's guidance as the source of truth.
• Follow anti-spam rules for commercial email. In the United States, the CAN-SPAM Act sets requirements for sender identification, subject lines, and unsubscribe handling. Other countries have different rules, and some are considerably stricter about prior consent. Verify the rules that apply to your recipients before sending.
• Apply extra scrutiny to regulated or high-risk fields such as medicine, law, finance, tax, and safety. Claims about returns, health outcomes, eligibility, or savings often need substantiation and approved wording. This guidance does not replace qualified professional judgment, approved procedures, or required authorization from a compliance reviewer, and you should not publish regulated claims without that review.
• Set up a simple approval path for every live change. Decide in advance who may pause a campaign, who may raise a budget above a threshold, and who signs off on new creative. Store approvals with the campaign record. A short approval log is often the first thing auditors and account managers ask for when something goes wrong.
• Monitor for brand safety and fraud. Watch for placements on sites that conflict with brand values, for click patterns that suggest bots, and for conversion spikes that do not match sales or lead records. Exclude suspicious sources and report anomalies to the platform when the tools allow it.

06Collaboration and documentation

• Campaign work involves marketers, designers, developers, sales teams, finance, and legal reviewers. Each group needs different information. Sales wants lead quality and response times, finance wants spend pacing and forecast accuracy, and developers need clear specifications for tracking events and landing page behavior. Agree on a single shared tracker so that everyone works from the same status.
• Keep a campaign log that records the objective, the hypothesis, the audience definitions, the bid and budget settings, the approval history, and every significant change. Write entries so that someone who joins the project next month can understand what happened and why. Include screenshots or exported reports for major decisions.
• Maintain a naming convention for campaigns, ad groups, ads, and tracking parameters. A consistent pattern, such as channel, objective, audience, and date code, makes reporting much faster. Avoid names that depend on one person's shorthand.
• Hold a short weekly review with the people who can act on the findings. Bring the dashboard, the change log, and two or three open questions. Keep the meeting focused on decisions, and assign an owner and a date to each action item.
• Document the tools and their settings. Record which ad platforms, tag management setup, analytics views, and reporting connectors are in use, and who has administrative access to each. Access control matters because one careless edit can change bids across many campaigns.

07Measures and continuous improvement

• Choose a small set of indicators tied to the objective. Common ones include cost per acquisition, return on ad spend, conversion rate from click to action, click-through rate on each ad variant, lead-to-sale rate, and average order value. Track trends over time rather than single-day swings, since daily volatility is normal in most accounts.
• Use confidence and sample size thoughtfully. Small numbers can look dramatic and mean very little. When a variant has only a few conversions, treat the difference as a hint rather than a verdict. Many marketers use a simple significance check, but the right threshold depends on the decision's cost and the business's tolerance for error.
• Measure incrementality when budget decisions are large. Platform-reported conversions often include people who would have bought anyway. Geographic holdout tests, paused-period comparisons, and controlled audience experiments can estimate how much lift the advertising actually created. These methods take planning and sometimes cost revenue during the test, so use them for decisions that justify the effort.
• Review results on a fixed cycle and feed them back into planning. A quarterly review might examine which channels met targets, which audiences held up, and which creative themes repeated. Update the brief template and the playbook with what was learned.
• Accept that the most useful improvements are often small and cumulative. A better landing page headline, a tighter negative keyword list, or a clearer call to action can outperform a dramatic change in bidding strategy. Keep a list of hypotheses ranked by expected impact and ease, and work through it steadily.

08Common mistakes and how to prevent them

• Optimizing for the wrong conversion event. When the ad platform optimizes for form starts rather than completed submissions, it finds people who start forms and abandon them. Prevention: define the event that matches real business value, and check it against downstream sales data every few weeks.
• Changing too many variables at once. A new audience, new creative, new bids, and new landing page launched together make results impossible to explain. Prevention: stage changes, or accept that the test is a bundle and say so in the report.
• Ignoring search term reports and placement reports. Broad match keywords and automated placements can spend heavily on irrelevant queries or sites. Prevention: review these reports on a schedule and add negatives or exclusions as part of normal maintenance.
• Letting budgets run without a cap or review. Automated campaigns can spend faster than expected when a bid strategy finds cheap inventory or when conversion tracking misfires. Prevention: set account-level alerts for spend thresholds and check pacing daily during launches.
• Reporting only the best-looking metric. A campaign with strong click-through rate may still have weak lead quality. Prevention: report the primary metric, the guardrails, and at least one downstream measure in every summary.
• Skipping the audit after a platform or website change. A redesigned checkout page or a new consent banner can break tracking silently. Prevention: test conversion paths after any site release and confirm that events still record correctly.

09Practical checklist

• Write a one-page brief with objective, primary metric, guardrails, budget range, and approval owner before building anything.
• Confirm that conversion events record once per real action and match downstream business records within a reasonable tolerance.
• Group keywords and audiences by intent, and align each ad group with one landing page and one core message.
• Check every ad and landing page against current platform policies and, for regulated claims, obtain qualified review.
• Verify consent and opt-out handling for any personal data used in targeting, and confirm the current rules with your privacy team.
• Launch with conservative budgets and avoid editing settings more often than the review schedule allows.
• Review search terms, placements, and audience segments on a fixed cadence and record every change with a reason.
• Report results with the hypothesis, sample size, attribution lens, guardrail metrics, and a clear recommendation.
• Store approvals, change logs, and reports in the shared campaign record so the work can be understood later.

★Learning sources

• Micro Degree: Digital Marketing and Online Advertising
• Micro Degree: Marketing Strategy Development
• Micro Degree: Programmatic Advertising Management
• Micro Degree: Lead Generation Operations