Octopus Energy's 'Magic Ink': AI that drafts support emails, with no job cuts
Octopus Energy's Kraken 'Magic Ink' drafted customer-service email replies with a human reviewing and sending each one. By end-April 2023 CEO Greg Jackson said it handled 34% of customer queries, 'the work of 250 people in the UK alone', at an 80% satisfaction rating versus 65% for humans, while the company said there would be no job cuts and announced thousands of new hires. The satisfaction figure varies across statements and is reported here with each source.
| Metric | Before | After |
|---|---|---|
| Share of customer queries handled by AI (end-April 2023) | human agents | 34%; stated as the work of 250 people in the UK |
| Customer satisfaction, AI vs human (May 2023 statement) | 65% human | 80% AI (figure varies across statements; see below) |
| Employment impact | human-run support | no job cuts stated; human reviews and sends every AI draft; 4,000 new roles announced in 2024 |
| Cumulative usage (Kraken case study, as-of date unspecified) | 6,239,087 calls summarised (695,379 hours); 9,415,901 messages generated |
The problem
Octopus Energy, a UK energy retailer, ran a high-volume customer-service operation dominated by email. In February 2023 it began trialling a generative-AI tool, ‘Magic Ink’, built into its Kraken platform, to draft replies to customer emails, with a human agent reviewing and sending each one (source). The interesting part is what did not happen next: no layoffs.
What was built
Magic Ink is Kraken’s generative-AI tool for customer service. Per Kraken’s own case study, it summarises interactions, drafts automated responses “as if written by the user”, and suggests actions, so human agents can focus on complex issues and phone calls (source). Every reply is reviewed and sent by a person; the system keeps humans in the loop in all cases (source).
The outcome
Share of queries and the “250 people” equivalence. Writing in The Times in May 2023 and quoted by Sifted, CEO Greg Jackson said the trial began “with a handful of customer emails with a person supervising it in February. By the end of April, it was answering 34% of all customer queries. That’s the work of 250 people in the UK alone, and it is doing it with an 80% satisfaction rating. Humans get 65%” (source). City AM relayed the same claim of “the work of 250 people” at an “80 per cent satisfaction rate, higher than the 65 per cent achieved by workers” (source), as did TheWrap (source). The “250 people” figure is a CEO estimate of workload equivalence, not a headcount cut.
The satisfaction figure is not stable across statements. This is reported, not merged. Kraken’s own case study puts it lower: “Around 35% of customer emails are currently written with the assistance of this tool, receiving higher customer satisfaction ratings (around 70%) than those without” (source). In a separate later talk Jackson gave the figures as 65% AI versus 55% human. Treat the 80/65 headline as one statement from May 2023, not a settled measurement.
No job cuts. Jackson said the technology was “unlikely to lead to job cuts” (source), and the company clarified that staff “supervise the answers AI provides”, drafting a personalised response “that a team member can review and then send on” (source). In 2024 Octopus announced 4,000 new jobs (source), making this an AI-plus-hiring story rather than an AI-for-layoffs one.
Cumulative volumes. Kraken’s case study reports that “around a third of messages generated by Magic Ink require zero to minimal changes before they are sent” and that, to date, it had “summarised 6,239,087 calls, the equivalent of 695,379 hours of talking time, and generated a further 9,415,901 messages” (source). The as-of date for these cumulative counts is not stated.
How this was verified
This case carries a green badge under TIN’s current standard: independently validated by TIN against the public record. It is unusually well-relayed for a lesser-known deployment: the headline is a CEO statement independently reported by Sifted, City AM and TheWrap, with a first-party Kraken case study adding operational detail and Octopus’s own hiring press corroborating the no-cuts framing. Read the badge as TIN validated this account is accurate, including the honest part: the figures are company-originated with no third-party audit, and the AI-vs-human satisfaction figure moves between statements, shown here and not merged. Direct client confirmation is the higher bar TIN grows into as the platform matures.
Sources
Cited in this case file. Tier 2 = independent press; Tier 3 = first-party or provider material. Each quote above was checked against an archived capture of the source.
- techUK, “Case study: Kraken Tech’s generative AI tool for customer service,” 2023 (Tier 3, first-party). https://www.techuk.org/resource/case-study-kraken-tech-s-generative-ai-tool-for-customer-service.html
- Sifted, “Octopus Energy CEO on AI answering customer emails,” 2023-05-09 (Tier 2). https://sifted.eu/articles/octopus-energy-ceo-chatgpt-news
- City AM, “AI doing the work of over 200 people at Octopus, chief executive says,” 2023-05-08 (Tier 2). https://www.cityam.com/ai-doing-the-work-of-over-200-people-at-octopus-chief-executive-says/
- TheWrap, “AI satisfies customers better than humans,” 2023-05-09 (Tier 2). https://www.thewrap.com/ai-satisfies-customers-better-than-humans/
- Octopus Energy, “Octopus Energy boosts energy tech revolution with 4,000 new jobs,” 2024 (Tier 3, first-party). https://octopus.energy/press/octopus-energy-4000-jobs/
Kraken 'Magic Ink'GPT-class language models
- Status
- verified
- Method
- Independently validated by The Internet Ninja against the public record: every figure below is quoted exactly from a cited independent or first-party source and checked against an archived capture, and on 2026-07-07 each load-bearing figure was re-verified against the live sources over the open network; all five cited sources were reachable and confirmed live. This is a public-record case validated by TIN, not a client-submitted engagement. The headline figures are CEO and company statements independently relayed by several outlets but not independently audited, and the AI-vs-human satisfaction figure is not stable across Octopus's own statements (80/65, around 70, and 65/55), shown here rather than merged. Direct client confirmation is the higher bar TIN grows into as the platform matures.
- Verified on
- 2026-07-07
- Provider
- Kraken Technologies (Octopus Energy Group), 'Magic Ink'
- Client
- Octopus Energy · Energy / utilities (UK)
- Disclosure
- named
