Website Marketing: Views, Location and Satisfaction

variables <- tribble(
  ~Variable,       ~Role,            ~Type,                ~Codes,
  "Gender",        "Control",        "Categorical",        "1 = Male, 2 = Female",
  "Views",         "Independent",    "Ordinal (segment)",  "under5, 5to10, 10to15, 15to20, over20",
  "Location",      "Independent",    "Categorical",        "1–7 (regions / stores)",
  "Revisit",       "Dependent",      "Metric",             "6–24 (Likert composite)",
  "Satisfaction",  "Dependent",      "Metric",             "6–24 (Likert composite)"
)

knitr::kable(variables, caption = "Variable definitions and coding scheme")

views_summary <- df %>%
  group_by(Views) %>%
  summarise(
    mean_revisit      = round(mean(Revisit, na.rm = TRUE), 1),
    mean_satisfaction  = round(mean(Satisfaction, na.rm = TRUE), 1),
    n                  = n(),
    .groups = "drop"
  )

knitr::kable(views_summary, caption = "Mean scores by view segment")

knit("
  Customers in the higher view segments (15to20 and over20) generally
  report higher Revisit and Satisfaction scores.
  - Under 5 views: average Revisit falls between 8–14.
  - Over 20 views: average Revisit climbs to 15–18.
")

ggplot(views_summary, aes(x = Views)) +
  geom_col(aes(y = mean_revisit, fill = "Revisit"),
           position = "dodge", width = 0.4, alpha = 0.85) +
  geom_col(aes(y = mean_satisfaction, fill = "Satisfaction"),
           position = position_dodge(width = 0.5), width = 0.4, alpha = 0.85) +
  scale_fill_manual(values = c("Revisit" = "#378ADD", "Satisfaction" = "#1D9E75")) +
  labs(title = "Mean revisit & satisfaction by view segment",
       y = "Mean score (6–24 scale)", x = "Weekly view segment", fill = NULL) +
  theme_minimal(base_size = 13) +
  theme(legend.position = "top")

location_summary <- df %>%
  group_by(Location) %>%
  summarise(
    mean_revisit      = round(mean(Revisit, na.rm = TRUE), 1),
    mean_satisfaction  = round(mean(Satisfaction, na.rm = TRUE), 1),
    .groups = "drop"
  ) %>%
  arrange(Location)

knitr::kable(location_summary, caption = "Mean scores by location")

knit("
  Location 7 stands out with consistently high Revisit scores (18–24)
  regardless of view segment. Location 1 lags behind, averaging 7–12
  on satisfaction. This suggests that site-specific factors (UX, content
  relevance, local marketing) moderate the views → satisfaction path.
")

gender_views <- df %>%
  group_by(Gender, Views) %>%
  summarise(mean_sat = round(mean(Satisfaction, na.rm = TRUE), 1),
            .groups = "drop") %>%
  mutate(Gender = ifelse(Gender == 1, "Male", "Female"))

gender_wide <- gender_views %>%
  pivot_wider(names_from = Gender, values_from = mean_sat)

knitr::kable(gender_wide, caption = "Mean satisfaction by gender and view segment")

knit("
  Both genders follow the same upward trend: satisfaction increases with view
  frequency. However, female respondents in the 5to10 segment show a slightly
  higher satisfaction peak compared to males, suggesting an engagement
  sweet-spot for that demographic.
")

mosaic <- df %>%
  mutate(high_revisit = Revisit >= 15) %>%
  filter(high_revisit) %>%
  group_by(Views, Gender) %>%
  summarise(count = n(), .groups = "drop") %>%
  mutate(Gender = ifelse(Gender == 1, "Male", "Female")) %>%
  pivot_wider(names_from = Gender, values_from = count, values_fill = 0) %>%
  mutate(Total = Male + Female)

knitr::kable(mosaic, caption = "High revisit counts (≥ 15) by segment and gender")

knit("
  Views per week is a stronger predictor of Revisit intention than Location.
  While specific locations (notably Location 7) perform exceptionally, the
  overarching trend is clear: increasing engagement frequency drives a
  near-linear increase in Satisfaction across both genders.

  Recommendation: Invest in strategies that move users from low-view segments
  into the 10to15+ bracket — the marginal gains in revisit and satisfaction
  are highest in this transition zone.
")