Conjoint analysis

Utilities from the raw choices, a simulator calibrated to real behavior with none and competitors in it, and price scenarios with share, revenue and source of volume

Customize this Agent

XLSXPPTX16 steps

The problem

Since September 23, Disney+ ad free costs $21.49, and $21.99 with Hulu

Read willingness to pay straight off the utilities and ad free looks safe above $21.49. Stated willingness to pay runs above real: a meta analysis of 77 studies found 21 percent on average, and more for indirect methods such as conjoint. A simulator not calibrated to the tier mix and the churn after the last increase, or missing none and competitors at new prices, sets January too high

The request

Fieldwork on Project Sorrel's pricing conjoint closed last week with 1,512 completes. We're setting January prices for our two streaming tiers, ad supported and ad free, 9 million US subscribers, after Disney+ and Hulu repriced September 23. Estimate HB utilities from the raw choices, price by tier, with the dual response none, and check fit and holdouts before anyone sees a share. Simulate with none and every competitor at today's list price, calibrated to our tier mix and the churn after our last increase. Run January grids for both tiers: share, subscribers, revenue including ads, and where lost subscribers go. Show stated against calibrated willingness to pay, and recommend prices in a readout deck. Files attached.

Sorrel CBC responses Oct26.csv
Sorrel CBC design.csv
Sorrel conjoint questionnaire v4.docx
Sorrel subscribers by tier Sep26.xlsx

How Axe works

ApprovedChanges asked

Your reviewApprove, or say what to change

Approved or changes

Simulator and readoutSimulator, deck and QC log

Final reviewApprove the revised files

Revised simulator and readoutRevised files and QC log

Step

Build the simulator

Prompt

Sent to the model at this step

Build the simulator in Excel on the checked utilities. Every scenario holds Sorrel's two tiers, every competitor plan from the price table as a concept on the design's attributes at its current list price, and none. Use randomized first choice with the scale factor tuned to the holdouts, so near identical plans, such as Disney+ and Hulu ad free at the same price, do not inflate their combined share, and show share of preference beside it. Weight respondents to the population targets in the design map. Then calibrate in two steps and record every factor. Mix: run today's prices and set one external effect per Sorrel tier so the split between Sorrel's tiers among those choosing Sorrel matches September's tier mix in the subscriber file. Price response: simulate Sorrel's last increase as it was made, with competitors at their prices on that date, and scale the price utilities by one factor until the predicted loss from the tier that rose matches its churn above the normal rate in the months after. Repeat both until both hold. Convert shares to subscribers by indexing each scenario to the base case on current subscribers by tier. Monthly revenue is subscribers times price for each tier, plus ad tier subscribers times ad revenue per ad tier subscriber. Derive willingness to pay by simulation, never by dividing part worths by the price coefficient: for ad free over the ad tier and for each feature, the price change that returns Sorrel's share to where it was, against every competitor and none, once on the uncalibrated simulator as stated and once calibrated.

Uses

Study
Subscriber data
Design map
Competitor prices
Checked utilities
Fit report

Produces

Simulator

What you get back

XLSXPricing simulatorUtilities, importances, fit, competitor prices, calibration, January grids and source of volume

PPTXReadout deckRecommended prices, share and revenue by scenario, stated against calibrated willingness to pay

QC logOpen items first, then every figure marked checked, corrected or open

Make it yours

Point it at your own study, swap in your simulator rule and calibration targets, or change the competitors you price against. Every step and every prompt can be edited

Customize this Agent

Opens a private copy of this Agent in your workspace, ready to adapt to your own sources

More agent examples

Market research analystsMarket forecastPPTXXLSX15 steps

Market research analystsMarket share analysisXLSXPPTX13 steps

Insights teamsTopline reportDOCXXLSX16 steps

Capital markets analystsBroker opinion of valuePPTXXLSX16 steps

Financial planning analystsBudget variance analysisXLSXDOCX13 steps