How Others Do It

Software Provider

Technology providers typically have complex distribution channels, selling direct over the web, through one, two and three tier distribution. Often data on actual sales can be spotty or even totally missing. So how can they determine whether their marketing is working?

Situation:  A large micro-site of a major software provider needed to determine whether the marketing to drive traffic to the micro-site was producing measurable increases in sales of software. The marketing for the micro-site was at the noise level against all of the other marketing being undertaken for the brand and so measuring anything directly was simply not possible. In addition, the site itself was not an e-commerce site, so conversions on the site were also not measurable. Lastly, the metrics infrastructure was home-grown and not terribly effective at delivering actionable metrics, although there was metrics to be had with the proper guidance. Marketing activities included email, e-newsletters, web-placement and content sharing, direct mail and trade shows.

Solution:  A project team was established, comprising internal marketing for the micro-site, the web metrics team and an outside online survey service.

1. Analysis of site visitors was done to classify visitors into high, medium and low value. High value – those that downloaded and voted on content; Medium value – those that downloaded only; low value – those that browsed only.
2. Using statistical modeling, marketing activities were correlated against time series to derive ROMI-factors for each type of visit value. For example, ROMI factors were expressed as cost in dollars for each 1,000 high value visitors.
3. Using online survey techniques, site visitors were asked to take a short survey asking them their change in likelihood to purchase and their likelihood to recommend following their visit to the site. Surveyed visitors were correlated back to their visit value type to determine differences in likelihood to purchase and recommend for high, medium and low value visits.
4. The results from points 4. and 5. above were combined to calculate the incremental purchase intent for each visit and hence for each marketing activity. This allowed a true end-to-end ROMI calculation of marketing investment to expected incremental purchase intent.
5. Purchase intent was verified against actual purchases through follow-up surveys 60 days after the initial survey to determine whether expressed purchase intent was a good predictor of future software purchase activity.

Results:  This analysis provided the marketing team invaluable answers to the question ‘what’s working and what’s not’. The marketing team was able to re-align its software tactics and focus on only those activities clearly driving incremental purchase intent (and recommendation). With this information, the team was able to realign over half its budget, doubling its investment in some areas and cutting off investment in others.

If you'd like to learn more about how you can apply this and other methods to improve your marketing activities, then join us at our next workshop.

To learn more, feel free to contact us at 770.570.8205 or if you have other questions you can email Susan.Howard @ MarketingROIWorkshop.com.