Section 2

Technical Questions

"You are a real estate agent, advertising real estate in the NorthEast US. You are setting up a campaign in Taboola to promote your business. How would you measure and how would you maximize your campaign's success?"

How I'd measure success

Running a real estate campaign through Taboola, I'd track:

Metric Formula Why it matters
CTR Clicks / Views Are people clicking? Low CTR means something needs adjusting.
CPC Cost / Clicks What am I paying per click?
Conversion Rate Leads / Clicks Are clicks turning into actual inquiries?
Cost Per Lead Spend / Leads The number that actually matters for budgeting.
ROAS Revenue / Spend Am I making money?

How I'd maximise performance

  1. Geo-targeting. Restrict to NorthEast US: Boston, NYC, Philadelphia, etc. Avoid paying for clicks from users outside the target area.
  2. Audience segmentation. First-time buyers, investors, relocators. Tailor messaging to each segment.
  3. Creative testing. Run multiple headlines and images to see what resonates. Test different angles: urgency, lifestyle, investment opportunity.
  4. Device performance. Monitor conversion rates by device. Allocate budget to higher performing devices.
  5. Publisher review. Analyse placement reports weekly. Exclude sites with poor conversion rates.
  6. Retargeting. Re-engage users who viewed listings but didn't enquire.
  7. Dayparting. Test different days and times to find when your audience is most active.

"What are the top 10 dates, based on total page views?"

SELECT
    date,
    SUM(desktop_views + tablet_views + mobile_views) AS total_views
FROM
    daily_views
GROUP BY
    date
ORDER BY
    total_views DESC
LIMIT 10;

This returns the date and total page views (combined from desktop, tablet, and mobile), grouped by date, sorted descending to show highest to lowest, and filtered to the top 10.

"Which 3 sites have the highest percentage of mobile traffic, out of their total traffic?"

SELECT
    s.id,
    s.name,
    SUM(dv.mobile_views) AS mobile_views,
    SUM(dv.desktop_views + dv.tablet_views + dv.mobile_views) AS total_views,
    ROUND(
        SUM(dv.mobile_views) * 100.0 /
        NULLIF(SUM(dv.desktop_views + dv.tablet_views + dv.mobile_views), 0),
        2
    ) AS mobile_pct
FROM
    sites s
INNER JOIN
    daily_views dv ON s.id = dv.site_id
GROUP BY
    s.id, s.name
ORDER BY
    mobile_pct DESC
LIMIT 3;
  • Joins sites and daily_views on site ID
  • Groups by site to combine all daily records
  • Sums mobile views and total views for each site
  • Calculates mobile as a percentage of total
  • NULLIF handles the edge case where total views is zero