Brands over time

Welsh Water
over time

Every year keeps its colour across every chart.│ ChatGPT, 30 Nov 2022

208 of Welsh Water's own blog posts, press releases and announcements, 2017 to 2026, measured sentence by sentence. How the writing has moved over the years, how it compares with 9 other water companies, and what it sounds like in its own words.

Pages measured
208
Years covered
2017–26
Archived before ChatGPT
0
Blog · press
106 · 102

01 · Homepage

One capture a year

Welsh Water's homepage from 2006 to 2023, closest to 1 July, from the Internet Archive. Pick a year or press play. Each headline is quoted from its capture and links to the archived copy.

2023dwrcymru.com30 Jun 2023
Welsh Water homepage, 30 Jun 2023

Screenshot of Welsh Water’s homepage (dwrcymru.com) as archived by the Internet Archive on 30 Jun 2023. © Welsh Water or its licensors. Shown for comparison and review; not endorsed by Welsh Water.

Homepage

2023

“Cost of living support”

Page title

Your not-for-profit water company | Dŵr Cymru Welsh Water

Open on the Internet Archive
  1. 2008How to contact us
  2. 2012Ask a Question?
  3. 2015Customer Experience Jobs
  4. 2017Check my water quality & hardness
  5. 2020For you Not for profit
  6. 2023Cost of living support

17 captures. Screenshots are rendered from the archive and can pull in images archived later than the page itself; dates come from the capture. Years without a capture are greyed out in the strip.

02 · Measures by year

How the writing moved

Every year’s pages, averaged. Pick a measure. Hatched bars rest on fewer than 10 pages: context, not a finding. The page count sits under each year.

Sentence lengthAverage words per sentence.
8
16
24
Water avg 16.0
16.0
18.4
19.5
20.0
18.3
18.3
18.3
18.8
17.4
16.8
ChatGPT · Nov 2022
2015no pages
2016no pages
20171 pp
201810 pp
201910 pp
202010 pp
202111 pp
202214 pp
202311 pp
20249 pp
202511 pp
202611 pp
10 or more pagesFewer than 10 pages: context, not a findingSector averageChatGPT released
Show as a table
YearSentence lengthPages
201716.01
201818.410
201919.510
202020.010
202118.311
202218.314
202318.311
202418.89
202517.411
202616.811

Years with pages: 2017 (1), 2018 (10), 2019 (10), 2020 (10), 2021 (11), 2022 (14), 2023 (11), 2024 (9), 2025 (11), 2026 (11). Dates come from page metadata, structured data, the URL, or the month a page was first archived (a latest possible date).

03 · Before and after

2019 to Nov 2022 against 2024 to 2026

The same measures on 45 pages from before ChatGPT launched and 31 pages from 2024 on. A change here is a change in style. It does not show how any page was written: people used every one of these words and shapes long before 2022, and SGR-011 found most of the shift across brands was a trend already under way.

2019 to Nov 2022 · 45 pp2024 to 2026 · 31 pp

Sentence length down 7%

words

  • 2019 to Nov 2022 · 45 pp
  • 2024 to 2026 · 31 pp

Source: SGR-011

Sentence length down 7%. words. Source: SGR-011
m2019 to Nov 2022 · 45 pp2024 to 2026 · 31 ppChange
Sentence length18.917.6−1.4

Sentence rhythm down 4%

How much sentence length varies (standard deviation over mean). Higher reads less even.

  • 2019 to Nov 2022 · 45 pp
  • 2024 to 2026 · 31 pp

Source: SGR-011

Sentence rhythm down 4%. How much sentence length varies (standard deviation over mean). Higher reads less even. Source: SGR-011
m2019 to Nov 2022 · 45 pp2024 to 2026 · 31 ppChange
Sentence rhythm0.740.71−0.03

Contractions up 76%

per 1,000 words

  • 2019 to Nov 2022 · 45 pp
  • 2024 to 2026 · 31 pp

Source: SGR-011

Contractions up 76%. per 1,000 words. Source: SGR-011
m2019 to Nov 2022 · 45 pp2024 to 2026 · 31 ppChange
Contractions6.711.7+5.1

"You" up 1 pt

share of I/we/you

  • 2019 to Nov 2022 · 45 pp
  • 2024 to 2026 · 31 pp

Source: SGR-011

"You" up 1 pt. share of I/we/you. Source: SGR-011
m2019 to Nov 2022 · 45 pp2024 to 2026 · 31 ppChange
"You"12%13%+1%

"We" down 8 pts

share of I/we/you

  • 2019 to Nov 2022 · 45 pp
  • 2024 to 2026 · 31 pp

Source: SGR-011

"We" down 8 pts. share of I/we/you. Source: SGR-011
m2019 to Nov 2022 · 45 pp2024 to 2026 · 31 ppChange
"We"81%73%−8%

Questions down 1%

per 1,000 words

  • 2019 to Nov 2022 · 45 pp
  • 2024 to 2026 · 31 pp

Source: SGR-011

Questions down 1%. per 1,000 words. Source: SGR-011
m2019 to Nov 2022 · 45 pp2024 to 2026 · 31 ppChange
Questions1.91.90.0

Exclamations down 4%

per 1,000 words

  • 2019 to Nov 2022 · 45 pp
  • 2024 to 2026 · 31 pp

Source: SGR-011

Exclamations down 4%. per 1,000 words. Source: SGR-011
m2019 to Nov 2022 · 45 pp2024 to 2026 · 31 ppChange
Exclamations0.230.22−0.01

Long words about the same

share of words

  • 2019 to Nov 2022 · 45 pp
  • 2024 to 2026 · 31 pp

Source: SGR-011

Long words about the same. share of words. Source: SGR-011
m2019 to Nov 2022 · 45 pp2024 to 2026 · 31 ppChange
Long words26%27%0%

Em dashes up 17%

per 1,000 words

  • 2019 to Nov 2022 · 45 pp
  • 2024 to 2026 · 31 pp

Source: SGR-011

Em dashes up 17%. per 1,000 words. Source: SGR-011
m2019 to Nov 2022 · 45 pp2024 to 2026 · 31 ppChange
Em dashes2.432.84+0.40

AI-associated words down 21%

per 1,000 words

  • 2019 to Nov 2022 · 45 pp
  • 2024 to 2026 · 31 pp

Source: SGR-011

AI-associated words down 21%. per 1,000 words. Source: SGR-011
m2019 to Nov 2022 · 45 pp2024 to 2026 · 31 ppChange
AI-associated words0.950.75−0.20

AI-era marketing words up 47%

per 1,000 words

  • 2019 to Nov 2022 · 45 pp
  • 2024 to 2026 · 31 pp

Source: SGR-011

AI-era marketing words up 47%. per 1,000 words. Source: SGR-011
m2019 to Nov 2022 · 45 pp2024 to 2026 · 31 ppChange
AI-era marketing words0.260.38+0.12

Quieter LLM-favoured words down 18%

per 1,000 words

  • 2019 to Nov 2022 · 45 pp
  • 2024 to 2026 · 31 pp

Source: SGR-011

Quieter LLM-favoured words down 18%. per 1,000 words. Source: SGR-011
m2019 to Nov 2022 · 45 pp2024 to 2026 · 31 ppChange
Quieter LLM-favoured words1.431.17−0.26

AI-associated sentence shapes up 9%

per 1,000 words

  • 2019 to Nov 2022 · 45 pp
  • 2024 to 2026 · 31 pp

Source: SGR-011

AI-associated sentence shapes up 9%. per 1,000 words. Source: SGR-011
m2019 to Nov 2022 · 45 pp2024 to 2026 · 31 ppChange
AI-associated sentence shapes0.430.47+0.04

Page length up 36%

words

  • 2019 to Nov 2022 · 45 pp
  • 2024 to 2026 · 31 pp

Source: SGR-011

Page length up 36%. words. Source: SGR-011
m2019 to Nov 2022 · 45 pp2024 to 2026 · 31 ppChange
Page length601818+217

Every chart is centred on its before value and shares one relative scale, so a longer bar means a bigger percentage change; the values themselves are in different units and do not compare. Before = 2019 to Nov 2022 (45 pages); after = 2024 to 2026 (31 pages). Method: SGR-011.

04 · Against its sector

Welsh Water among water companies

Each dot is one brand’s mean across all its measured pages; the tick is the sector average. Only brands with 10 or more measured pages are shown or counted. Select a dot to open that brand.

  1. Questions

    145% more than its sector

    3.4 against a sector average of 1.4; highest of 10.

  2. Em dashes

    35% more than its sector

    2.61 against a sector average of 1.93; 3rd highest of 10.

  3. "You"

    16% less than its sector

    20% against a sector average of 24%; mid-table, 5th of 10.

Sentence lengthMid-table, 5th of 10
Welsh Water 16.8
10.720.8
ContractionsMid-table, 5th of 10
Welsh Water 11.3
6.817.1
"You"Mid-table, 5th of 10
Welsh Water 20%
3%50%
"We"4th lowest of 10
Welsh Water 65%
40%95%
QuestionsHighest of 10
Welsh Water 3.4
0.03.4
Long words4th lowest of 10
Welsh Water 25%
24%31%
Em dashes3rd highest of 10
Welsh Water 2.61
0.563.24
AI-associated wordsMid-table, 6th of 10
Welsh Water 0.93
0.532.20
ExclamationsMid-table, 5th of 10
Welsh Water 1.16
0.072.62
Page lengthHighest of 10
Welsh Water 1,356
5141,356

Pages measured: Thames Water 106 · Severn Trent 115 · United Utilities 153 · Anglian Water 26 · Wessex Water 97 · South West Water 185 · Welsh Water 208 · Affinity Water 100 · Scottish Water 16 · South East Water 232. Measures and method: SGR-010; this sector and its brands were added by Brands over time, measured exactly as the studies measured theirs. Sector averages and ranks cover every brand in the sector on this site.

05 · Vocabulary

Words Welsh Water uses more than its peers

The words this brand reaches for more often than the rest of its sector. This is subject matter more than voice.

interest56.4 vs 3.2
financial58.4 vs 9.9
finance22.3 vs 2.7
wales24.7 vs 0.9
payments22.1 vs 1.0
cash19.5 vs 1.5
Welsh WaterSector peers

Uses per 10,000 words, ranked by weighted log-odds (Monroe et al. 2008) across 208 pages. Checked for product names and page furniture.

06 · In its own words

What it sounded like

Openings of real pages, quoted word for word, oldest first. Each links to the page, or to the Internet Archive’s copy when it has gone.

2019 · Press
“Welsh Water, the only not-for-profit water company in Wales and England, announced the ambitious target as part of its plan to increase the number of customers who interact with it in the language.”
Target to quadruple Welsh users launched by Welsh Water

07 · Every page

All 210 pages, by year

The full list behind every number above, with links to the live page or the archived copy. 208 count towards the measures; the rest were not measured (2 under 150 words of text). Pages with no reliable date count in the overall numbers but in no year.

202611 pages
202511 pages
20249 pages
202311 pages
202214 pages
202111 pages
202010 pages
201910 pages
201810 pages
20171 page
In the overall numbers, no year110 pages
Not measured2 pages

08 · Method

How this was measured

The pages

Welsh Water's own blog posts, press releases and announcements, fetched from the live site in 2026 and from the Internet Archive for 2015 to November 2022. Homepages are shown as captures; product and help pages are not measured.

The measures

Code counts from ScriptGrain’s writing instruments: sentence length and rhythm, contractions, pronouns, punctuation, long words, and three published lists of AI-associated words. No AI model reads or judges any page.

What it cannot say

These measures describe style. They cannot tell who or what wrote a page, and nothing here says or implies that any brand used AI to write.

Sample sizes

Every number carries its page count. Years under 10 pages, and before-and-after comparisons with under 10 pages on a side, are shown as context only.

Sentence length
Average words per sentence.
Sentence rhythm
How much sentence length varies (standard deviation over mean). Higher reads less even.
Contractions
Contractions (we're, you'll, it's) per 1,000 words. A rough gauge of how conversational the writing is.
"You"
Share of I/we/you pronouns that are "you": how much the writing addresses the reader.
"We"
Share of I/we/you pronouns that are "we": how much the writing talks about the organisation.
Questions
Question marks per 1,000 words.
Exclamations
Exclamation marks per 1,000 words.
Long words
Share of words with seven or more letters.
Em dashes
Em dashes per 1,000 words.
AI-associated words
Words from SGR-007's three AI-associated lists, per 1,000 words. These are style markers that people used long before ChatGPT; a higher rate says nothing about who or what wrote the page.
AI-era marketing words
Marketing words that became more common in AI-era copy (seamless, elevate, unlock...), per 1,000 words. A style measure only.
Quieter LLM-favoured words
Everyday words language models are known to over-use, per 1,000 words. A style measure only.
AI-associated sentence shapes
Sentence patterns such as "not just X, but Y", found by ScriptGrain's cliche checker, per 1,000 words. A style measure only.
Page length
Average words per page.

Studies: SGR-009: Do brands in the same sector sound the same? · SGR-010: How UK brands write · SGR-011: Did brand writing change after ChatGPT?.