Your First Week With Alesta

Use your first week with Alesta to correct the baseline, not to execute every recommendation. Review identity first, then source scope, performance and SEO evidence, competitors, and generated documents. End the week with one approved decision card, a list of missing evidence, and owners for corrections.

Before day one

Complete the domain baseline preparation guide. You should know the canonical domain, primary decision, important markets, review owners, and current product constraints.

The current free onboarding baseline is one-time and bounded. Plan the review while the evidence is fresh.

Day 1: Confirm identity and scope

Review company identity

Check:

• canonical domain;
• company and product name;
• visible description;
• apparent audience;
• primary offer and CTA;
• product category or frame;
• important source or page failures.

Mark each item:

• confirmed;
• correct in product, wrong on website;
• wrong in extraction;
• disputed internally;
• unavailable;
• needs a named owner.

Do not proceed as if disputed identity were correct. Generated competitor and strategy conclusions can inherit it.

Record the evidence boundary

Write down:

• pages and roles represented in the sample;
• pages or states outside scope;
• mobile and desktop coverage;
• field-data availability;
• provider failures;
• retrieval timing.

Day 1 deliverable

An approved or correction-assigned identity card plus a one-paragraph scope note.

Day 2: Read measurements without blending them

Start with real-user field data

When Chrome UX Report data is available, note:

• URL or origin scope;
• mobile or desktop;
• LCP, INP, and CLS status;
• 75th-percentile interpretation;
• reporting period and retrieval date.

When it is unavailable, write "unavailable." Do not substitute zero or a lab score.

Then review lab diagnostics

For mobile and desktop PageSpeed evidence, note:

• exact tested URL;
• test profile and time where available;
• major diagnostic clues;
• whether the result reproduces;
• which template or state may be affected.

Do not turn an automated opportunity into an engineering requirement before tracing the cause.

Day 2 deliverable

A short performance note that distinguishes user evidence, diagnostic evidence, scope, and the next investigation.

Use the performance-data guide if field and lab results conflict.

Day 3: Review SEO and public page evidence

Check the dependency order

Review important findings in this order:

1. discovery and valid response;
2. rendering;
3. intended indexability;
4. canonical consistency;
5. internal links and architecture;
6. page purpose, titles, headings, and content;
7. structured data where relevant;
8. experience and accessibility.

Confirm sample reach

For each material finding, ask:

• Which exact URL was observed?
• Is the page canonical and intended for indexing?
• Does the condition repeat by template?
• Was the result observed or inferred?
• What Search Console, crawl, code, or log evidence is needed next?
• What decision changes if confirmed?

Reject noise

Mark stale, expected, irrelevant, or non-reproducible findings with a reason. An honest rejected finding improves the backlog.

Day 3 deliverable

No more than five validated candidates split into correct, investigate, implement, accept, or reject.

Day 4: Correct competitors and positioning context

Classify proposed competitors

Use:

• direct rival;
• adjacent rival;
• substitute;
• status quo;
• search rival;
• aspirational peer;
• exclude.

Ask sales and customer-facing people which alternatives appear in real decisions. Add manual processes and services where relevant.

Review the public position

Check whether the generated context accurately represents:

• priority audience;
• trigger and job;
• category;
• product mechanism;
• value;
• proof;
• meaningful alternative;
• limitations and availability.

Correct product facts before editing persuasive language.

Day 4 deliverable

A validated competitor set and a list of product or positioning corrections with owners.

Use the competitor validation guide for a deeper process.

Day 5: Review documents and choose one decision

Read generated documents last

Now review product, competitor, and strategy documents against the corrected evidence.

Highlight every material statement as:

• supported;
• plausible inference;
• unsupported;
• contradicted;
• stale or conditionally available;
• requires private evidence.

Choose one action or research decision

Create a decision card:

Decision:
Evidence and scope:
Why now:
Unknowns:
Response: correct, investigate, test, implement, accept, or reject
Owner:
Required outcome or learning:
Acceptance evidence:
Guardrail:
Review date:
Stop or revise condition:

Day 5 deliverable

One approved card, a correction queue, and a missing-evidence request.

The end-of-week evidence packet

Your packet should contain:

1. corrected identity card;
2. scope and availability note;
3. field-versus-lab performance interpretation;
4. validated technical and on-page candidates;
5. adjudicated competitor set;
6. corrected product and positioning context;
7. one owned decision card;
8. a list of private sources needed next.

What not to do in the first week

• Do not publish generated strategy without review.
• Do not ask engineering to fix every warning.
• Do not treat unavailable data as zero.
• Do not present a bounded crawl as full coverage.
• Do not accept similar-looking companies as confirmed competitors.
• Do not promise a ranking, conversion, or revenue outcome.
• Do not describe llms.txt as a Google ranking tool.
• Do not assume free onboarding includes backlinks or observed AI mentions.
• Do not imply that the baseline automatically refreshes.

A practical review meeting agenda

10 minutes: Identity and scope

Confirm company, product, audience, domain, sample, and unavailable sources.

15 minutes: Measurements

Review field and lab data separately, then validate technical candidates.

15 minutes: Competitors and position

Approve roles, remove false candidates, and correct product claims.

15 minutes: Strategy review

Challenge high-impact recommendations and request missing evidence.

5 minutes: Decision

Assign one card, owner, and review date.

Troubleshooting

Identity is wrong

Check whether the public website is ambiguous. Correct the underlying workspace context and assign public-site corrections if needed before trusting strategy.

Field performance is unavailable

Keep the state unavailable. Use lab diagnostics and another valid real-user source if the organization has one.

Proposed competitors look wrong

Classify and remove them. Add alternatives from customer and sales evidence. Review positioning documents after the set is corrected.

The strategy sounds generic

Check whether the business decision, audience, private evidence, and constraints were available. Replace generic recommendations with sourced decision cards.

There are too many findings

Group root causes, validate valuable scope, and choose a small portfolio. Use the audit action-plan article.

What comes after week one

Extend the baseline with the private evidence the decision requires. The AI marketing audit guide covers customer, commercial, product, channel, and measurement layers. The fractional CMO scenario shows a 30-day extension.