Case study P01

From $80K to $455K MRR, with search traffic up 9x and CAC down 60%

Growth was not sustainable and churn was eating it. The fix started at the top of the funnel, with a programmatic search play that won spots on the results page we would not normally have ranked for.

Lumen5 logoLumen5Sr. Growth Marketing Manager & AdvisorDec 2019 – Nov 20223 min read
$80K$455K
Monthly recurring revenue
80K750K+
Monthly search sessions in 12 months
−60%+
Customer acquisition cost
7-fig
ARR growth in year one

Context

Lumen5 is an AI video creation platform that turns written content into video. About 65% of the business came from enterprise clients and 35% from self-serve, single-user accounts.

When I joined, growth was not sustainable. Churn ran as high as 33% in bad months, which put a hard ceiling on how big the business could get. My goal was to add $300K in monthly recurring revenue by significantly increasing the number of users, while bringing churn down to lift that ceiling.

I reported to the VP Marketing for the first half of my tenure. When they left, I was promoted to lead the marketing team, reporting to the CEO.

The thesis

Most teams fight for the ten blue links. The bigger opportunity is often everywhere else on the results page.

Competing for the whole results page

Result typeWhat it isHow we competed
InformationalHow-to and explainer resultsProgrammatic pages built around high-volume topics
TransactionalResults for people ready to actPages that went straight to making a video
LocalResults tied to a placeLocalized long-tail pages generated for location-based searches
Rich and thumbnail resultsVisual results with images or previewsSchema markup that made our pages eligible for these spots

A search results page is not one list. Depending on the query, Google mixes transactional results, informational results, local results and rich results with thumbnails, such as products and recipes. Each of those has its own rules for what gets shown.

We used structured data (schema) to make our pages eligible for the visual, thumbnail-style spots that video and template pages would not normally win, which put Lumen5 in front of searchers in places competitors were not competing.

The programmatic search engine

01 →Find demandTerms with 1,000+ monthly searches
02 →Generate the pageFrom a tested template
03 →Add schemaEligible for rich results
04 →Go localLocation pages for the long tail
05MeasureRankings, traffic, sign-ups
Winning terms and templates decided what we built next

The engine ran on simple rules. Any search term with more than 1,000 monthly searches got its own page, generated from a template. If the term could also be paired with a location, the system went one step further and created localized pages for those longer-tail searches, which supported the main page and captured demand competitors ignored.

Every page carried structured data, so it could compete for rich results as well as standard listings. Results fed back into which terms and templates we expanded next.

Layering in paid

Once search was carrying the base, we added a large paid program across Google, Meta and native ad platforms. Paid was there to scale what was already working, not to carry growth on its own.

Organic traffic changed the economics of paid. As search brought in more customers at almost no marginal cost, blended acquisition cost fell by more than 60% even while monthly paid spend rose by six figures.

Results

Monthly search sessionsWithin 12 months
Before80K
After750K+
Monthly recurring revenueOver my tenure, against a +$300K goal
Before$80K
After$455K

Monthly search sessions grew from about 80,000 to more than 750,000 within 12 months, and the channels I ran drove seven-figure ARR growth in that first year. Over my full tenure, monthly recurring revenue grew from $80K to $455K, beating the $300K goal.