Client name withheld under NDA.
The retailer sold bouquets, potted plants and gift sets, mostly through its website, with same-day delivery across Singapore. Business peaked twice a year, around Valentine's Day and Mother's Day, and the rest of the year was quieter. Social media was run in bursts by whoever in the team had time. Posts went up before the peaks and then thinned out, comments on photos waited days for a reply, and nobody could say what social had actually earned.
We started by agreeing on one definition of return, because the word gets used loosely. In this case study, return means revenue attributed to social divided by the total cost of social. Total cost is the management fee plus every dollar of media spend, so nothing is left out. Before changing any posts, we measured the baseline: S$4,000 of attributed revenue a month against S$2,500 of total cost, which is a return of 1.6x.
Before any new posting, we set up the tracking. Every link in a bio, story or post carried its own tracking tag. We created a promo code for each platform and added a single question at checkout asking how the customer heard about the shop. We also reviewed the past year of posts to see which ones had drawn orders, not just likes.
Without tracking, a return figure is a guess. Agreeing on how revenue would be credited in month one meant the later numbers could be trusted and compared fairly with the baseline.
We built a calendar with a steady weekly rhythm instead of bursts. It mixed behind-the-scenes florist work, new arrivals, gift ideas by budget and customer stories. We wrote a short voice guide so every caption sounded like the shop. We also set same-day delivery cut-off times clearly on posts, so nobody was disappointed.
A steady rhythm keeps the shop in people's minds between occasions. Gift buyers rarely plan far ahead, so they tend to buy from the shop they saw last week.
For Valentine's Day we planned three weeks of content in advance: early ordering reminders, bundle ideas, last-order dates and delivery slot updates. We put a small paid boost behind the posts that were already earning orders, and no budget behind posts that were not. Community replies were answered within the same working day.
Peak days are won in the weeks before. Boosting only proven posts kept media spend small, and quick replies turned questions about delivery into completed orders.
We replied to every comment and direct message, and added a saved reply set for common questions about delivery areas, card messages and substitutions. We invited past buyers to share photos of their flowers and featured the best ones. A short reminder post went out ahead of birthdays and anniversaries that customers had mentioned.
Many gift buyers have the same few dates every year. Treating them as regulars, not one-off orders, is where the rise in repeat customers came from.
We ran the same planning approach for Mother's Day, adjusted for what we learned in February. Reporting each month showed attributed revenue, total cost, return, orders and average order value, with the checkout survey results alongside the promo code and link data. We also checked the numbers by month so the peaks were visible.
Applying the lessons from the first peak shows whether the approach holds. Reporting by month, not as one blended figure, kept the picture honest about which months were boosted by the season.
By month 6, the return on social had moved from 1.6x to 4.1x. Monthly attributed revenue rose from S$4,000 to S$14,365, while total cost rose from S$2,500 to S$3,500 as we added a modest media budget. Orders from social grew from 50 to 169 a month, and average order value rose from S$80 to S$85. Month 6 included Mother's Day, so it was a peak month. In the four months outside the two peaks, the return averaged 3.2x, which is the fairer figure for a normal month.
Tracking tags, promo codes and checkout survey live, weekly calendar and voice guide in place.
Valentine's Day build-up, same-day replies, saved answers and customer photo features running.
Reminder posts for repeat dates, Mother's Day campaign and monthly return reporting.
The 4.1x is attributed revenue divided by the total cost of social, fee and media together. The baseline of 1.6x used the same definition. When someone quotes a multiple, ask what is in the top line and what is in the cost.
Month 6 included Mother's Day and reached 4.1x, but the four non-peak months averaged 3.2x. Both are good, and the second is what to plan around. Ask any agency to show quiet months next to peak ones.
Tracked links, promo codes and a checkout survey capture people who clicked, used a code or told us. They miss someone who saw a post and searched for the shop later. The true return is probably a little higher than 4.1x, though promo codes can also be shared, so we treat the figure with care.
How we plan, post and manage social accounts for Singapore businesses.
Photos, short videos and posts made for each platform.
Reaching local shoppers on Facebook and Instagram.
A homeware retailer, where we measured results in orders.
Tell us about your shop and how customers find you. We will set out how to measure social properly and what a realistic six-month plan looks like.