Using audience lifetime value to guide social testing


Goal: treat audience attention as an economic signal: connect social behavior to lifetime value, then allocate acquisition effort toward the cohorts that stay and spend.
ChefSteps built its audience around ambitious cooking videos, recipes, and professional kitchen equipment. Its Facebook page had reached 12,000 likes, but the company wanted to know which followers were becoming customers in an effort to value their social media community and the time they spent there. After an ambitious CLV quantification attempt failed to achieve statistical relevance, we pivoted to analyzing content, audience behavior, and advertising performance, finding directional success with a Facebook test promoting handmade Japanese knives that produced 3.7× return on ad spend among existing fans.



Valuing a fan
Ahead of a product launch, ChefSteps wanted to understand how its social audiences related to customer value. Crown Social Agency analyzed audience cohorts and tested creative and targeting approaches, linking channel decisions to a lifetime-value model rather than treating every interaction as equivalent.
Situation
ChefSteps was an early-stage, venture-backed culinary education company founded by Modernist Cuisine alumni, and had built a fast-growing community of chefs and food enthusiasts. Strong content engagement did not answer the questions leadership wanted to solve before acquisition planning. ChefSteps wanted to know which social touchpoints contributed to sales, how acquisition and retention should change when viewed through customer lifetime value, and where high-performing audiences could be reached at or below a target acquisition cost.
Scope
The work had to connect behavior across owned, earned, and paid channels rather than credit the last visible click. It also had to distinguish a large or active audience from a valuable one, then turn that diagnosis into repeatable media decisions.
Analysis
CROWN proposed a two-phase diagnostic: customer-lifetime-value and content analysis first, followed by acquisition optimization. The team mapped user journeys and engagement patterns across the digital ecosystem, correlated sign-ups, social engagement, and purchase data, and used social listening to locate high-value conversations. The resulting matrix put individual content and channel performance on top of the same denominator.
Execution
The acquisition phase tested four classes of social advertising: external website ads, channel-growth ads, content-engagement ads, and sponsored stories using social proof. Sixty-four creative variants combined four headlines, four body-copy options, and four images. Targeting focused on male chefs ages 25 to 44, while iterative copy and audience analysis connected the experiments back to the channel-level value model.
Evidence and measurement
The operating question was not which channel produced the most clicks, but which touchpoints brought in customers who stayed and spent. CROWN compared customer lifetime value across communities and authentication paths, then looked for acquisition opportunities below the target cost.
Outcome
The analysis identified Facebook-authenticated users as the highest-value audience cohort. Segmentation added additional improvement of click-through rate and sales efficiency on Facebook, and the team established a repeatable method for channel-level lifetime-value modeling; however, incomplete identity data, unreliable connections between acquisition sources and purchases, and sample size prevented us from developing a statistically defensible model.
Hindsight
Our original goal of segmenting customers based on additional online behaviors was ambitious, especially so inside of a startup with a limited number of fans and customers to segment.
Crown Social pivoted to working with ChefSteps to improve end-to-end ecommerce tracking and refocusing the analysis on measurable sales, audience behavior, and advertising efficiency, establishing a clearer picture of content performance and early acquisition economics and providing recommendations for subsequent advertising and publishing decisions. We also found directional success with a Facebook test promoting handmade Japanese knives that produced 3.7× return on ad spend among existing fans (vs 0.9x ROAS for friends of fans) across combinations of four headlines, four body-copy variations, and four images.
Captions
Ahead of a product launch,
ChefSteps wanted to understand how its
social audiences related to customer value.
Crown Social Agency analyzed audience cohorts
and tested creative
and targeting approaches,
linking channel decisions to a lifetime-value
model rather than treating every interaction
as equivalent.
Audience cohorts
and creative tests connected social
decisions with a customer lifetime-value
model ahead of a product launch.
Different interactions can carry different value.
The model helped frame what to test. After an ambitious CLV quantification attempt failed to achieve statistical relevance, we pivoted to analyzing content, audience behavior, and advertising performance. The strongest response came from existing fans. They generated a 4.834% click-through rate and 3.7× ROAS. Friends of fans returned 0.90×, while two cold-audience segments generated no recorded sales.




