How the growth engine works
Seven stages, from a search to a customer who keeps sending.
The previous version described three jobs: rank, get quoted, convert. Those remain, but they are the middle of the engine, not the whole of it. Discovery comes before ranking, activation and measurement come after conversion, and learning is what makes the second quarter better than the first.
Discover
Which questions, from which people, at which moment?
Discovery is done before a title is written. The unit of work is a question a Canva user asks after creating something, not a keyword with volume. Each candidate passes six tests: is it searched by a Canva user, is Constant Contact the logical next step, can it enter the answer without being forced, does it lead to trial or integration use, does it strengthen the ecosystem relationship, and is the traffic commercially meaningful rather than just large. Seventeen topics from the previous plan failed at least three of those tests and were removed.
Query validation before publication
Every primary and secondary query is checked in keyword tooling and Search Console. Volumes are reported as ranges. Nothing on the strategy page is stated more precisely than the data supports.
Checked against what exists
Every URL is checked against the live blog, product pages, knowledge base and community before it goes on the calendar. Existing pages become anchors; they are refreshed and linked, not competed with.
One question per page
The question is chosen first. The URL, title, opening passage and headings follow from it. No two pages target neighboring queries.
AI question intent recorded
For each page, the version of the question a person would type into an assistant is written down alongside the search query, because the two are often phrased differently and the page must answer both.
Rank
Hold a position on the query, and on the group around it.
Constant Contact's domain authority is not the constraint; coverage and page depth are. The blog captures fewer ranking terms per URL than its nearest competitor, which means the fastest gains come from pages that answer a whole question group completely rather than from more pages. Each article is built to capture the primary query, its rephrasings and the long tail beneath it.
Hub-and-spoke, never sideways
Articles link up to the hub and to their cluster anchor. No sibling links. Depth accrues to the anchor, and no two pages pull against each other.
Substance over length
Every article carries a comparison table, a worked mini-case, a mistakes checklist and a long-tail FAQ. Word count is a by-product, not a target.
Technical baseline
Clean URL, one H1, question-shaped H2s from real query data, descriptive image alt text, internal links with descriptive anchors, published and modified dates set, canonical set, Core Web Vitals within the template's existing performance.
Refresh cadence
Pages that rank but do not convert, or convert but slip, are refreshed in month four and quarterly after. Integration facts are re-verified at every refresh.
Get cited
Be the passage an answer engine lifts, and the entity it names.
A page can rank well and never be cited, because retrieval systems need passages that stand alone, can be verified, and name their entities in full. Google AI Overviews, ChatGPT, Gemini, Perplexity and Copilot are each sampled monthly on a fixed question set. No article is promised a citation on a timeline; what is promised is that every article is built to earn one.
The answer goes first
The opening paragraph answers the title question in 40 to 60 words and is complete on its own. That passage is the one most likely to be lifted.
Per-page retrieval specification
Each article specifies its search intent, AI question intent, extractable answer, primary entity, supporting entities, evidence required, conversion action and internal-link destination before drafting. The roadmap carries the summary; the brief carries the full version.
Entities named in full
"Constant Contact" and "Canva" are written out, never "the platform" or "the tool". Product capabilities are described in first-party terms with the source they come from.
Evidence, not adjectives
Every claim about the integration links to the product page, the knowledge base or the launch release. Statistics carry the date they were true and a source. No claim is made that cannot be checked.
Structured data
BlogPosting and BreadcrumbList on every article; HowTo where the page is a procedure; FAQPage where the FAQ is substantive. Eight FAQs base, expanded to thirteen on the substance pass. Validated before publish.
Clear authorship and dates
Byline "Constant Contact Editorial Team" unless the marketing team assigns another, with visible published and updated dates. Answer engines weight both.
Convert
Turn the reader into a trial, or a connected account.
A reader who arrives, finds the answer and leaves has cost money and returned a session. Every article carries one conversion block, written for one reader, placed where the problem has just been established, and pointed at one destination that matches the intent. Sending every article to the same trial button is the mistake this version corrects.
CTA matched to intent
Informational and how-to pages route to the hub or the integration guide. Comparison and decision pages route to the trial. Workflow and template pages route to the connect step and the import action, because much of that traffic is already a customer.
One offer per page
One block, one destination, one measurable click. Anchor pages carry a second placement lower down. Nothing else competes for the click.
Trial framed by the moment
Where the trial is the destination, the block says what happens next: connect Canva, import the design, send. No card required is stated once, plainly, because it is true and it removes the last objection.
Tagged at source
Every CTA carries the article, cluster and intent group in its parameters so trial starts can be attributed by content, not just by channel.
Activate
Connect Canva, import a design, send a campaign.
A trial that never connects Canva is a trial the program did not really produce. Activation is the event the program is ultimately reported against: a Canva account connected, an asset imported, a campaign created with Canva creative, and a campaign sent. NativeCode cannot see these events; the product team can. The methodology here is about making them more likely and making them visible.
Content that continues inside the product
How-to articles are written so that the steps match the in-product flow exactly, including the names of buttons and menus, so a reader who starts a trial mid-article can finish the task without re-orienting.
Instrumentation
The program tracks five product events for the content-sourced cohort: connected, imported, created, sent, published to social. Time to first campaign is derived from them.
Existing customers count
A large share of Canva-related traffic will already be a Constant Contact customer. Connect and import actions from that audience are reported as activation too, because they are the same workflow overlap the partnership exists to grow.
Retention content
The measure-stage article and the vertical workflows are written for the second campaign, not the first, because the second campaign is where retention is decided.
Measure
Report the funnel, not the traffic.
The five-level hierarchy on the strategy page is the reporting structure. Levels 1 and 2 are NativeCode's from month one. Level 3 uses the signup attribution already in place. Levels 4 and 5 draw on product data and are reported jointly with the analytics team. Keywords per page remains a useful diagnostic and appears in the appendix of each report, not in the headline.
Monthly report, one page first
Page one is the funnel: impressions, clicks, CTA clicks, trials, and whatever activation data is available. Diagnostics follow. Executives read page one.
Fixed AI question set
Twenty questions, sampled across five engines on the same day each month, recording whether Constant Contact is named, cited, or absent. Reported as a share, with the questions published so the method is auditable.
Cohort, not aggregate
Canva-related content traffic is a defined cohort in analytics. Its trial rate, activation rate and retention are compared to the site baseline, not reported alone.
Honest timelines
Indexation and first impressions in two to four weeks; long-tail positions and first citations in four to eight; movement on primary terms in eight to sixteen; first-page positions on the lowest-difficulty targets and attributable trials in sixteen to twenty-four. Stated as expectations, not guarantees.
Learn
Make the next month better than this one.
Every month closes with three decisions: which pages to refresh, which CTAs to rewrite, and which queries the data has surfaced that the calendar did not anticipate. December's optimization phase applies this across the whole set and produces the first quarter plan on validated demand rather than assumption.
Refresh what ranks but is not cited
Usually a passage problem: the answer is present but not self-contained. Fixed by rewriting the opening and the FAQ, not by adding length.
Rewrite what converts below the cohort
Usually a CTA mismatch: the wrong destination for the intent. Fixed by changing the destination before changing the copy.
Add what the data asked for
Search Console surfaces the questions people actually typed. Those become the January calendar, ahead of anything still on paper.
The Index as the learning asset
The proposed Small Business Marketing Index turns the connected-account base into published evidence, if the legal, privacy and technical basis exists. It is scoped in December and does not appear in any report before it is real.
Every article, with its intent, query, URL, CTA and measurement goal.
See the roadmap →