Is digital marketing worth it before you spend on it?
01 — Overview
Bamboo Holiday Lighting installs, maintains, removes and stores exterior holiday lighting across South Florida. Two years old, seasonal, growing entirely through door-to-door sales, and running no social accounts at all. The question was whether digital marketing would move the business enough to justify the spend — and if so, which channels.
02 — Scope
Decision problem
How would digital marketing and online operations affect business activity?
Research problemWhich digital channels are cost-effective for a seasonal business?
Research problemWhich kinds of social advertising are most effective and most trusted?
Direct access to the business meant the research ran on real operating conditions rather than assumptions.
03 — Method
01
US holiday lighting valued at $8.09B in 2024, projected to $10.86B by 2032. Purchases are habitual and seasonal — customers tend to stay with the same provider year over year, which makes first acquisition disproportionately valuable. Local competitive set identified and compared on pricing, durability, response time and installation quality.
02
Chosen over focus groups deliberately: in group settings participants soften or shift answers toward the room. One-on-one removes that and allows probing into the why behind an answer.
03
Screened on three criteria: South Florida resident, active social media user, able to purchase holiday lighting. 134 responses collected, 100 qualified. Every non-qualifying response was excluded rather than adjusted.
04
Coded to a purpose-built codebook, then univariate analysis across all variables and multivariate testing of three stated hypotheses by chi-square.
Every non-qualifying response was excluded rather than adjusted.
| # | Question | Type | What it was for |
|---|---|---|---|
| 1 | Are you currently living in south Florida? | Dichotomous | Screening |
| 2 | Do you own or rent your current residence? | Nominal | Housing status |
| 2b | Which best describes your living situation? | Nominal | Conditional branch |
| 3 | Do you typically decorate your home for the holidays? | Nominal | Behavior |
| 3b | Even though you don't currently decorate, would you be interested in putting up home or holiday lighting in the future? | Nominal | Conditional branch |
| 4 | Do you actively use social media platforms? | Dichotomous | Screening |
| 4b | Even though you don't actively use social media, do you still browse or see posts from platforms like Facebook, Instagram, or TikTok? | Nominal | Conditional branch |
| 5 | Have you ever purchased home/holiday lighting? | Nominal | Screening · history |
| 6 | When discovering new local businesses, where are you most likely to come across them first? | Categorical | Finding 01 · 03 · H2 |
| 7 | What factor most influences your trust in a holiday lighting company found online? | Categorical | Finding 03 |
| 8 | Which platform do you find most trustworthy for evaluating a company's reputation? | Categorical | Finding 03 · H2 |
| 9 | How much does a company's social media presence affect your decision to contact or purchase from them? | Likert 1–5 | H1 |
| 10 | Which types of social media content would make you most interested in home/holiday lighting? | Select top 2 | Finding 04 · H1 |
| 11 | What type of promotion would most likely encourage you to purchase holiday lighting services? | Categorical | Finding 05 · H3 |
| 12 | Describe a specific instance when you saw a social media post or ad that made you want to contact or purchase from a company. | Open-ended | Coded, 7 categories |
| 13 | What specifically stood out about that post or ad that influenced your interest? | Open-ended | Coded, 7 categories |
| 14 | How likely are you to share or tag a company's lighting post if you find it visually appealing or inspiring? | Likert 1–5 | Behavior |
| 15 | How old are you? | Ordinal | Demographics |
| 16 | What is your gender identity? | Nominal | Demographics |
| Variable | Description | Coding |
|---|---|---|
| LivingSituation | Own or rent current residence | 1 Own · 2 Rent · 3 Other |
| Decor | Typically decorates for the holidays | 0 No · 1 Yes · 2 Maybe |
| SMedia | Actively uses social media platforms | 0 No · 1 Yes |
| LightingHistory | Has purchased home/holiday lighting before | 1 Within the past year · 2 More than a year ago · 3 No, but interested |
| BusDiscovery | Where new local businesses are first discovered | 1 Instagram · 2 TikTok · 3 Facebook · 4 Google search · 5 Word of mouth · 6 Other |
| TrustFactor | Factor most influencing trust in a company found online | 1 Online reviews · 2 Word-of-mouth · 3 Real videos/photos · 4 Professional appearance · 5 Other |
| TrustPlatform | Most trustworthy platform for evaluating reputation | 1 Facebook · 2 Google · 3 Instagram · 4 TikTok · 5 Snapchat · 6 Other |
| SMediaPres | How much a company's social presence affects the decision | 1–5, not at all → a lot |
| SMC_Photo … SMC_HowTo | Preferred social content type, one flag per option | 0 No · 1 Yes |
| PROMO | Promotion type most likely to encourage purchase | 1 Percentage discount · 2 BOGO · 3 Referral · 4 Free consultation · 5 Loyalty reward |
| OPENENDED1 / 2 | Motivating example, and what stood out about it | Coded to 7 categories each |
| SHARE | Likelihood of sharing or tagging a lighting post | 1–5, extremely unlikely → extremely likely |
| AGE / GENDER | Demographics | AGE 1–6 bands · GENDER 1 male · 2 female |
04 — Findings
Where Bamboo was
54%
discover local businesses on a social platform — Facebook, Instagram and TikTok combined
34%
through Google Search
69%
decorate for the holidays
66%
of those who don't said they'd like to
Bamboo had no accounts and no search presence. A market almost entirely reachable through channels the business was not present on.
Chi-square test of discovery platform against most-trusted platform
χ² = 32.90, df = 20, p = 0.04
The one significant result in the study.
Consumers trust the same platform they use to find businesses. Someone who discovers a contractor through Google trusts Google Reviews; someone who finds them on Facebook trusts what Facebook shows them.
The practical consequence: presence has to be consistent across platforms rather than concentrated on the cheapest one, because credibility doesn't transfer between them.
Google: first to be found, and first to be trusted
Google is both the leading discovery channel and the most trusted platform, at 59%. Instagram and TikTok are meaningful among younger and renting respondents.
54
of 100 put installation and results videos in their top two
Interviews said the same thing and explained why: before-and-after footage lets a customer see the process and the result on a house like theirs.
Percentage discount 37 · BOGO 22
Descriptively the strongest drivers of booking intent, though the hypothesis testing this did not reach significance.
05 — Hypothesis testing
H1
p = 0.84
χ² = 4.17, df = 8
Not significant
Video and testimonials appeal broadly regardless of how active someone is online
H2
p = 0.04
χ² = 32.90, df = 20
Significant
Consistent credibility signals needed on every platform, not just one
H3
p = 0.32
χ² = 9.23, df = 8
Not significant
Descriptively, percentage discounts and BOGO lead; the effect isn't statistically supported
Two of three hypotheses failed. Both are reported here for the same reason they're reported in the paper: a result that doesn't hold is information, and dropping it would make the one that did hold look stronger than it is.
06 — Recommendations
Before-and-after photos and installation video as the primary content. Facebook community groups for local visibility.
Search optimization around South Florida holiday lighting terms, and actively soliciting reviews and customer project photos from satisfied clients.
Google is both the leading discovery channel and the most trusted.
Pre-season percentage discounts and referral rewards, timed ahead of the November planning spike.
Target: South Florida homeowners aged 25–54 — the largest and most active segment in the data.
07 — Notes
The sample skewed young and social-active.
That aligns with the digital marketing question but under-represents Bamboo's older homeowner customers, which means the results are probably more optimistic about digital than a broader age mix would be.
No HOA representation.
Not one board member or property manager in the sample. In this region a single HOA decision can affect dozens of homes at once and can restrict which contractors enter a community — so the study says nothing useful about the largest available account type.
Response bias was visible in the data.
Answers clustered toward the top of the scale across unrelated questions, which suggests anchoring rather than independent evaluation. Some respondents also likely overstated their support for local business.
Convenience sample, off-season timing.
Participants were recruited by accessibility rather than randomly, and the data was collected months before the November window when people actually start thinking about holiday lighting.
Given the project again, the first thing I'd change is recruiting HOA decision-makers deliberately rather than hoping they turn up in a general sample.