LikeFolio Consumer Demand
A nowcast of what consumers are buying, wanting and complaining about — before it shows up in revenue.
We measure the brands, products, and trends people actually talk about and search for, roll them up to the company (or its individual divisions), and track how that demand changes week to week. It comes from two independent sources — X and Google search — read at two levels: individual companies, and economy-wide consumer trends.
Coverage
History from Jan 2021
History from 2016
Brand demand — two sources
| X (social behavior) | Google search interest | |
|---|---|---|
| What it measures | What people say and intend about brands | What brands people search for |
| Metrics | 16 behavior metrics (below) | 1: search interest |
| Brands tracked | 9,508 brands & products mapped to companies | |
| Served at | company / division | company / brand |
| Frequency | daily | daily |
| History start | Jan 2021 | Sep 2020 |
| Point-in-time | yes | yes |
| Status | production | final review |
Both sources are built by tracking thousands of individual brands. On X we aggregate them up to the division (and company); on Google search each brand is available as its own series, and also rolls up to the company.
Consumer trends — two sources
| X (social behavior) | Google search interest | |
|---|---|---|
| What it measures | Which trends people discuss taking part in | Which trends match consumer web searches |
| Metrics | 1: trend intent | 1: search interest |
| Mapping | ~1,450 trend-to-company links · ~165 trend-to-dataset links | |
| Frequency | daily | daily |
| History start | 2016 | 2016 |
| Status | production | final review |
The two primary sources, X and Google Trends, are largely independent — one captures conversation, the other captures intent to find — so read together they cover more of real demand than either alone. Search interest is placed on a single comparable scale across all brands and dates so any two can be compared directly.
The 16 metrics (X data)
Every public post about a covered brand is classified into the behaviors below. Illustrative examples.
| Metric | Signals | Example post |
|---|---|---|
| Mentions | any brand reference (the base volume) | "grabbing a Celsius before the gym" |
| Purchase intent | planning / about to buy | "think I'm finally buying the new iPhone this weekend" |
| Negative intent | planning to quit / cancel | "that's it, I'm cancelling Netflix" |
| Positive sentiment | happy with the product | "my new Hokas are the comfiest shoes ever" |
| Negative sentiment | unhappy with the product | "the new Chipotle burrito is disgusting" |
| Switch in | moving to the brand | "transferred to Robinhood — so much better" |
| Switch out | moving away from the brand | "switched from Dropbox to OneDrive today" (counts as OUT for Dropbox, IN for OneDrive) |
| Promotion | deals, sales, sweepstakes | "Chipotle has free delivery today" |
| Price — value | feels like a good deal | "TJMaxx has some legit deals, wow" |
| Price — pain | feels too expensive | "$15 value meal at McDonald's is ridiculous" |
| Defect | product broken / faulty | "iPad screen died after six months" |
| Outage | service down | "is Snapchat down? can't log in" |
| Availability | sold out / can't get it | "2nd place I've been to that's sold out of Modelo" |
| Safety | a safety concern | "the new Ulta concealer gave me a rash, anyone else?" |
| Labor | strikes, walkouts, staffing | "short staffed at Olive Garden again, manager said it's a problem" |
| Political | boycotts, political reaction | "boycotting Nike after that ad" |
Cleaned before you see it
Bots, spam and giveaway farms are removed in our pipeline — they never appear in your data. Beyond that, every remaining volume spike is detected and reviewed (over 14,000 reviewed to date) and categorized. You choose which categories to include; by default only genuine consumer Demand is counted.
| Spike type | What it is | In the default series? |
|---|---|---|
| Demand event | Real consumers wanting, using or buying — a launch, a drop, something going viral | Included |
| Marketing | Company-driven campaigns, sponsorships, sweepstakes | Optional |
| Incident | Recall, outage, lawsuit, controversy | Optional |
| Investor | Earnings, M&A, analyst and stock chatter | Optional |
| Meme | The brand name used as a joke or template, unrelated to the product | Optional |
Turn any category on to include its spikes, or leave it off to exclude them — your call, per metric. The default (Demand only) is the cleanest read of genuine consumer demand. The spike-review method is proprietary.
How you get it
- Delivery: REST API and bulk Parquet / S3; query by company id, ticker, FIGI or permID.
- Point-in-time: every series carries a knowledge date, so you can accurately backtest.
- What ships: derived counts only — never post text, ids or raw third-party payloads.
- Trial: a six-month-lagged feed with full history. Request access →