Dolly Parton search hit 100 while storm Dolly also peaked

A live Google Trends board is not a single story. At 2026-09-02T10:01:36Z, get_top_trends ranked dolly parton 1, tim curry 2, tropical storm dolly 11, dolly parton cause of death 8, and did dolly parton die 17. get_growth on Google Search for the week of 2026-08-29 returned score 100.0 for dolly parton against a 7D baseline of 1.0, growth +9,900 percent. The same week tropical storm dolly also scored 100.0, but every window other than 14D baselined at 0.0 and returned growth 999999 with a low-quality warning. A job that stores one row named "dolly" mixes a celebrity spike, a rumor-shaped query, and a storm. Split the keywords before the write.

What did the live boards show on 2026-09-02?

get_top_trends with type Google Trends, limit 25, as_of 2026-09-02T10:01:36Z:

The 7D rank_change sort on the same board returned one mover against baseline 2026-08-26T02:01:36Z: lake ontario from prev_rank 139 to rank 5, rank_change +134. dolly parton was not in that mover list. Pair board rank with get_growth. Absence from rank_change means the token was already high enough on 2026-08-26 that a 7D rank delta did not fire, not that Search was quiet.

Wikipedia Trending at the same timestamp put Dolly Parton at rank 5. Google News Top News, limit 15, had no Dolly token in the top 15 (lead rows were geopolitics and a Nepal flood count). Reddit keyword dolly parton returned 404 no_data. A monitor that only reads the news board misses the Search spike. A monitor that only reads Search still has to split celebrity, rumor, and storm intents.

Source fields for the Search board are on Google Trends. Wikipedia board fields are on Wikipedia Trends. The loop that stores board rank plus growth is on Demand sensing API.

How far above the historical series is dolly parton search?

get_time_series on Google Search, keyword dolly parton, weekly points. Prior peaks in this pull never cleared 7.

The 2026-08-29 week is the series peak at 100. Almost every other week in the five-year pull sat at 0 or 1. get_growth windows vs that 100.0 tip:

Those percent figures are real point-to-point math against a 1.0 baseline. They are not a quality warning. The finding is the 100 vs 7 gap against every prior peak, not the 9,900 percent label.

Did YouTube, News, and Wikipedia move with Search?

Yes on YouTube and News. Wikipedia moved on a monthly grain. Reddit did not return a series.

Google Search dolly parton, week 2026-08-29, score 100.0 vs 1.0 on 2026-08-22, 7D +9,900 percent.

YouTube dolly parton scored 100.0 vs 2.0 on 2026-08-22, 7D +4,900 percent. The 14D, 30D, 3M, and 12M YouTube windows were the same 100.0 vs 2.0. Video interest moved with Search. That is a usable 7D feature.

Google News dolly parton scored 100.0 vs 1.0 on 2026-08-22, 7D +9,900 percent. 14D, 30D, 3M, and 12M News windows baselined at 0.0 and returned growth 999999. Print 100.0 vs 0.0. Do not store 999999. The live Google News Top News board still omitted the name from the top 15 at 2026-09-02T10:01:36Z, so keyword News coverage and the Top News board are different feeds.

Wikipedia is monthly. Keyword Dolly Parton on source wikipedia used recent_date 2026-08-01, score 100.0, volume 22,170 page views. 7D and 14D returned collapsed_range (not enough history for those presets). 30D vs 2026-07-01 (2.7 / 674) was +3,603.7 percent score and +3,189.32 percent volume. 3M vs 2026-05-01 (3.6 / 856) was +2,677.78 percent. 12M vs 2025-08-01 (3.3 / 798) was +2,930.3 percent. Do not join the August Wikipedia month onto Search week 2026-08-29 without storing both dates.

How is tropical storm dolly different from dolly parton?

Call both keywords. Do not substring-match dolly.

POST https://api.trendsapi.ai/api
Authorization: Bearer <api_key>
Content-Type: application/json

{"mode":"get_growth","source":"google search","keyword":"tropical storm dolly","percent_growth":["7D","14D","30D"]}

import json
import os
import urllib.request

API = "https://api.trendsapi.ai/api"


def get_growth(keyword: str, source: str) -> dict:
    payload = {
        "mode": "get_growth",
        "source": source,
        "keyword": keyword,
        "percent_growth": ["7D", "14D", "30D", "3M", "12M"],
    }
    key = os.environ["TRENDSAPI_API_KEY"]
    req = urllib.request.Request(
        API,
        data=json.dumps(payload).encode("utf-8"),
        headers={
            "Authorization": f"Bearer {key}",
            "Content-Type": "application/json",
        },
        method="POST",
    )
    with urllib.request.urlopen(req, timeout=60) as resp:
        envelope = json.loads(resp.read().decode("utf-8"))
    if envelope.get("statusCode") != 200:
        raise RuntimeError(envelope)
    # body is a JSON string; parse it a second time
    return json.loads(envelope["body"])


parton = get_growth("dolly parton", "google search")
storm = get_growth("tropical storm dolly", "google search")

tropical storm dolly Search scored 100.0 on 2026-08-29. 7D, 30D, 3M, and 12M baselined at 0.0 and returned growth 999999. The 14D window is the only one with a non-zero baseline: 100.0 vs 1.0 on 2026-08-15, growth +9,900 percent. The payload also set data_quality to low with the warning that over 90 percent of the series is zeros. Print 100.0 vs 0.0. Do not treat 999999 as a feature.

YouTube tropical storm dolly scored 74.0 vs 0.0 on every requested window, again growth 999999 and the same low-quality warning. Celebrity YouTube is 100.0 vs 2.0. Storm YouTube is 74.0 vs 0.0. Those are two series, not one "dolly" row.

Google News tropical storm dolly scored 100.0 vs 0.0 on every window, 999999, low-quality. Keyword News moved. The Top News board still did not list the storm in the top 15.

Related Search tokens on the live board moved with the celebrity name, not the storm. carl dean scored 100.0 vs 0.0 on 7D (999999) and 100.0 vs 1.0 on 14D (+9,900 percent). dolly parton cause of death and did dolly parton die both scored 100.0 vs 0.0 on every window, 999999, low-quality. Those queries are search tokens on the board. They are not a confirmation of an event. Store them as separate keywords with the low-quality flag. tim curry (live rank 2) also scored 100.0 vs 0.0 on 7D. A celebrity cluster on the board is not evidence that every name shares a cause.

What about lake ontario, the only 7D climber?

lake ontario is the control. Live rank 5, prev_rank 139, rank_change +134, the only 7D mover. Search week 2026-08-29 scored 92.0 vs 7.0 on 2026-08-22, 7D +1,214.29 percent. 14D, 30D, and 3M were 92.0 vs 9.0, +922.22 percent. 12M was 92.0 vs 7.0 on 2025-08-30, +1,214.29 percent. The baseline is 7.0 or 9.0, not 0.0, so the percent is usable. A rank_change job would have caught lake ontario and missed dolly parton. A get_growth job on a watchlist that already contained dolly parton would have caught the 1.0 to 100.0 jump and missed lake ontario unless that token was on the list. Run both.

How should a scheduled job store this?

One POST per keyword per source. Parse body a second time. Persist keyword, source, recent_date, recent_value, baseline_value, growth, and the data_quality flag when present. Drop or quarantine rows where growth equals 999999. Keep the raw scores.

A Monday Dagster asset that already watches orchestrator packages can add a small newsroom watchlist of exact phrases (dolly parton, tropical storm dolly, lake ontario) without changing the HTTP contract. That wiring is on Dagster Trends API. The thin newsroom page is newsroom Trends API.

Quota math: 3 keywords × 2 sources (google search, google news) is 6 get_growth calls, plus 2 get_top_trends boards (Google Trends, Wikipedia Trending) is 8 calls. Weekly Monday cron is about 32 successful 200s in a 30-day month. Free tier is 100. Daily polling of that set is 8 × 30 = 240, which needs Starter (5,000) or higher. Only HTTP 200 responses count. A 404 no_data (Reddit here) does not.

Do not collapse dolly parton and tropical storm dolly because they share a token. Do not treat a rumor-shaped query at rank 8 as ground truth. Do not skip rank_change because the rank-1 token is already famous. The 2026-08-29 week is the first time dolly parton Search left single digits in this series. Storm Dolly peaked the same week on a near-zero history. Those are two writes.