Work / Crystal Art Gallery (CAG)
B2B + B2C ยท Brand and growth
Crystal Art Gallery (CAG): a 10-brand art and decor company
The problem
Fine Art Canvas, the company's main e-commerce brand, had four goals: a lower cost per purchase, a higher conversion rate, a better return on ad spend and more repeat customers. Mobile was 73 to 80% of traffic. Customers also could not see what the frame colors and textures really looked like, and the unboxing felt flat: no tissue, no special touch, and extra-large canvases were hard to open.
What I did
- Conversion work. I rebuilt the home and product pages and tested changes with Microsoft Clarity A/B tests. The winners were a new add-to-cart button color, a cleaner nav bar, reviews higher on the page, and no sticky add-to-cart bar. I also added a static home banner, value and trust icons, a features accordion on mobile, and a larger search bar with better filters.
- Product and unboxing. I reshot the frame colors and textures and added a custom packing label.
- Paid creative. I directed creative for paid social with a vetted agency: testimonials, Scott Living concepts, catalog ads and a lifestyle video.
- Partnership. I was the lead marketer on the HGTV Scott Brothers partnership. I worked with their team on scripts, shipped the photo-shoot assets on time, and the work landed in their magazine, on their landing pages, website and social channels.
- Measurement. I used multi-touch attribution to move budget toward what worked.
- Philanthropy. A project with HealthRIGHT360 and Prototypes Philanthropy. I directed and produced a gifted video for Prototypes, a sub-agency of HealthRIGHT 360. See the film.
The result
Net salesAd spend
View as a table
| Week | Net sales | Ad spend |
|---|---|---|
| 10/2 - 10/8 | $20,168 | $6,251 |
| 10/9 - 10/15 | $16,095 | $5,732 |
| 10/16 - 10/22 | $15,747 | $5,612 |
| 10/23 - 10/29 | $17,402 | $5,955 |
| 10/30 - 11/5 | $17,600 | $6,048 |
| 11/6 - 11/12 | $16,791 | $5,994 |
| 11/13 - 11/19 | $18,395 | $5,570 |
| 11/20 - 11/26 | $20,271 | $5,627 |
| 11/27 - 12/3 | $23,515 | $6,318 |
| 12/4 - 12/10 | $18,676 | $5,815 |
| 12/11 - 12/17 | $18,088 | $5,726 |
| 12/18 - 12/24 | $16,217 | $5,430 |
| 12/25 - 12/31 | $17,380 | $5,570 |
| 1/1 - 1/7 | $21,011 | $6,092 |
| 1/8 - 1/14 | $17,291 | $6,074 |
| 1/15 - 1/21 | $21,634 | $6,525 |
| 1/22 - 1/28 | $21,748 | $6,544 |
| 1/29 - 2/4 | $27,189 | $5,946 |
| 2/5 - 2/11 | $19,143 | $5,337 |
| 2/12 - 2/18 | $14,137 | $5,330 |
| 2/19 - 2/25 | $17,209 | $6,958 |
| 2/26 - 3/4 | $17,512 | $6,208 |
| 3/5 - 3/11 | $24,028 | $8,298 |
| 3/12 - 3/18 | $17,467 | $6,162 |
Source: Agency weekly ad report. Net sales are all Shopify store sales, not only ad-attributed.
Over the 24 weeks, ad spend stayed in a tight band of about $5,300 to $8,300 a week while weekly net sales ranged from $14,137 to $27,189. The best week, January 29 to February 4, returned $4.57 in sales for every ad dollar. The week of November 27 returned $3.72. The weakest, February 12 to 18, returned $2.65. Across all 24 weeks the return was $3.13 per ad dollar.
How this was measured: from my own pitch deck: $15M in year one, then +160% to $39M. I led digital and content strategy for both sides. The growth belongs to the whole company and team.
How this was measured: from the agency's weekly report. The sales figure is all store sales, not only ad-attributed. Platform ROAS was 2.78 and cost per purchase $46.32.
How this was measured: from the agency's creative report for one five-week period.
How this was measured: from the partnership deck. For context, one 2026 agency benchmark treats 8% to 15% monthly net list growth as a healthy target (Darkroom, Email Benchmarks 2026). That is a target, not an industry average. My +421% came within about six months. At that benchmark, six months of growth compounds to roughly 59% to 131%.
How I got there
Weekly reviews of Google Analytics and the agency's creative reports decided what to test. I shipped one change at a time, kept the winners from the A/B tests, and moved budget toward the ads and pages that earned it.
Stack and workflow
What I learned
The biggest lesson was that A/B testing never finishes. With this many ads sending traffic, the product page did the real selling, so we kept testing product pages and treated UX and UI as part of the media plan. An ad can only earn what the page behind it converts, so the page had to keep evolving too.
Reference
"...a great communicator, a creative problem solver and a strong project manager."
Krisztina worked with Tim on the HealthRIGHT360 and Prototypes Philanthropy project.