Marketing Analytics: Attribution modeling
Download as PPTX, PDF0 likes476 views
The document outlines a analysis of marketing channels using attribution modeling, highlighting the effectiveness of paid and organic search in generating revenue. It details customer purchase latency, indicating that while some customers purchase immediately, others may take over 12 days. Recommendations for reinvestment suggest increasing budgets for organic and paid search channels to optimize marketing efficiency, while emphasizing the importance of other channels for branding purposes.
Marketing Analytics: Attribution modeling
- 1. Attribution Modeling Marketing Analytics 310 Fall 2015 IzmirVodinaj
- 2. Outline ▪ Paid Search Analysis: What touch point should be given more credit? ▪ Organic SearchAnalysis: What is organic search best at? ▪ Customer Purchase Latency: How long does it take a customer to purchase? ▪ Channel Investment Analysis: Where should the most money be invested in?
- 3. Paid Search: Attributed vs. Last Click Model Attributed Revenue Last Click Revenue $91,648.76 $75,461.15 91648.7563 75461.15 Paid Search Channel: Attributed vs. Last Click Model Attributed Last Click The paid search revenue calculation through the attributed model exceeds the revenue through the last click model calculation. This means that more credit is given to multiple touch points during the conversion than to the last click. Question #1 Insight
- 4. Organic Search Purchase Path Analysis Introducer Influencer Closer 36 65 229 229 65 36 0 50 100 150 200 250 Organic Search Purchase Path Closer Influencer Introducer The organic purchase path analysis shows that customers are generally informed and ready to purchase when they buy through the organic search due to the high number of closers. Question #2 Insight
- 5. Customer Purchase Latency Analysis Mean Mode Median 12.1 0 1 12.1 0 0 0 2 4 6 8 10 12 14 Mean Mode Median Customer Purchase Latency Customers make decisions of purchase instantly and buy the product the same day. However, some customers take over 12 days after the first visit. Finally, this data is very helpful for marketing managers to determine inventory and make predictions for future sales. Insight Question #3
- 6. Marketing Investment Decisions: ROAS Based on Channels Direct Email Organic Search Paid Search Social Attributed 3.48 1.33 6.6 8.33 0.41 Last Click 3.17 1.23 8.82 6.86 0.51 Direct, 3.48 Email, 1.33 Organic Search, 6.6 Paid Search, 8.33 Social, 0.41 3.17 1.23 8.82 6.86 0.510 1 2 3 4 5 6 7 8 9 10 DollaronAdSpend Return on Ad Spend Organic Search and Paid Search generate the highest marketing efficiency. In attribution view, paid search scores highest while in last click view organic search brings the highest return. Insight Based on the analysis laid out above, the retailer should spend the invest another $5,000 in Organic Search and Paid Search Marketing Channels Recommendation Question #4
- 7. Marketing Channel Re-Investment Decisions Direct, 3.48 Email, 1.33 Organic Search, 6.6 Paid Search, 8.33 Social, 0.41 3.17 1.23 8.82 6.86 0.510 1 2 3 4 5 6 7 8 9 10 Direct Email Organic Search Paid Search Social DollaronAdSpend Return on Ad Spend Bonus Investment/Channel Direct Email Organic Search Paid Search Social Original investment 30000 20000 12500 11000 1000 Recommended Investment 18500 4000 26000 25000 1000 Looking at the data, the biggest re-investment focus should be organic search and paid search, but one must not neglect the other channels because they serve branding purposes the return of which may not easily be calculated the same way as our current calculation. Recommendation
Editor's Notes
- #4: Pivot Table: Filter: Marketing Channels Paid Search Values: Sum Revenue Attributed, Sum Revenue Last Click
- #5: Pivot table: Filter Marketing Channel, Organic Search Values: Position Type
- #6: Steps: Copy visit date and order date in another sheet. Then adjust the dates by min and max. After that, find the latency by subtracting the visit date from the order date. Finally, find mean, mode, and median respectively. =AVERAGE(G:G), =MODE.SNGL(G:G), =MEDIAN(G:G)
- #7: Steps: Pivot Table Rows: Marketing Channels Values: Revenues Last Click, Revenues Attributed Division of revenue over investment
- #8: Steps: Pivot Table Rows: Marketing Channels Values: Revenues Last Click, Revenues Attributed Division of revenue over investment