Tell Me More

Welcome!

Learn more about the Textmetrics text optimizing tool:

• Optimize Written Text
• Language Translation & Interface
• Randstad HR Services case study
• BCR & ROI back-of-the-envelope calculations
• Technical information regarding the AI that powers the tool

To see how Textmetrics could work in your organization:

Text Optimizer

Textmetrics advocates using generative AI but believes users need need more control over AI output. That’s why we offer software that might seem to be generative AI, but offers a lot more. Textmetrics’ Text Optimizer checks the generated text for several content-related issues, such as readability, credibility, any bias. Texmetrics’ software is GDPR-compliant and doesn’t learn from user input and data.

• Writing tool analyzes text for reading complexity, DEI issues, formality, sophistication, style guide and preferences.
• Generates improvements and assists staff in writing high-quality content. It reads and assesses while the content is being written and offers suggestions to improve the text.
• Algorithms recognize content characteristics such as readability for intended audience, how friendly, credible, neutral, feminine, masculine, etc.

Text Optimizer Features

• Readability – CEFR, ILR, ACTFL
• DEI module from Extanto
• Gender message balancing
• Accessibility WCAG standards
• Customizable to clients’ standards (corporate guidelines, publishers standards, learning objectives)

AI Function Augmentation

Textmetrics can be used to enhance AI functionality. It is a comprehensive AI ruleset governance tool, addressing issues with accuracy, credibility, bias, DEI, language validation and copyright sometimes included in AI results. Its advanced algorithms and linguistic insights provide a robust solution to enhance content quality and reliability. Simply put, it is quality control for AI-generated responses. Importantly, the integration model gives reliable structure to AI business logic and predictability to how those results are rendered for the UI / end user: 

• AI: LLM creates content, foundational element, processing and generating human-like content interactions.
• Platform Interaction: UI provides an intuitive interface for clients to engage with the language model. 
• Control: interprets user inputs, leveraging NLP to understand context and intent, and guides the LLM to generate coherent, contextually relevant and accurate responses. 

Language Analysis & Translation

• Textmetrics translates into forty languages
• Textmetrics performs language analysis in the following twelve languages: Danish, Dutch, English, French, German, Italian, Japanese, Norwegian, Polish, Portuguese, Spanish and Swedish 
• Textmetrics will also display the user interface in any of the languages in which analysis is performed

Language Validation

• Grammar & Spelling: Rules for distinct regional linguistic dialects such as Canadian French (Quebecois). 
• Plain Language: A government standard making it easier for the public to
  read, understand, and use government communications. It is, “writing that is clear, concise, well-organized, and follows other best practices appropriate to the subject or field and intended audience.” (US Plain Writing Act of 2010) 
  • Find what they need
  •  Understand what they find the first time they read or hear it
  • Use what they find to meet their needs 
• Inclusivity Guidelines: Incorporation of organization-specific directions for multiculturally sensitive communications. The goal is to provide guidance that includes everyone regardless of background, sexual orientation, gender expression, religion, ethnicity, race or ability. 
• Organization style guide: Rules to standardize the authoring, formatting, designing and publishing written communications. The US government uses the style manual created by the Government Publishing Office (GPO).

Reading Profiency

Readability indices determine content suitability for knowledge base target audiences. Textmetrics uses the ILR, CEFR or ACTFL guidelines to help authors communicate at specific audience reading levels. These guidelines describe language comprehension abilities informing the writer of the suitability of vocabulary and phrasing for the intended audience. Textmetrics makes recommendations that will either elevate or simplify the content to optimize the audience’s understanding of the content’s meaning.

• Interagency Language Roundtable (ILR) – used in the USA for federal agencies.
• Common European Framework of Reference (CEFR) – used across Europe.
• American Council on the Teaching of Foreign Languages (ACTFL) – typically used in business and academia in the USA.

HR Job Description Case Study

Randstad
Job description must communicate core values clearly:

• Use language & important elements to attract right candidates
• Make descriptions simple
• Make job description easily found on internet
• Must match Randstad branding

This is a complex collection of language factors hard to apply correctly 100% of the time.

Textmetrics analyzes all elements that will increase conversion:

• Gives recruiter suggestions to maximize success of JD
• Uniform voice- AI uses corporate branding and marketing guidelines in suggestions
• Trains recruiters how to write effective job descriptions
• Attracts more diverse, higher quality field of candidates

Randstad recognized a 30% increase in efficiency using Textmetrics augmented writing tools.

Back-of-the-envelope calculations for BCR & ROI

Assumptions regarding our calculations:

• Size of company: 1000 employees
• # of HR employees (assuming 1.4 HR employees per 100 employees): 14 employees
• HR annual Costs: $1.3M (avg. salary of $85,000 each + benefits)
• Recruitment Costs (~15% of HR annual costs) = $196k
• Time Savings have been a range of 14-30 % for past clients. For our purposes we’ll use 22%.
• Time Savings(~22% of Recruitment Costs) = $43k

Benefit Cost Ratio/ Return On Investment

For a company with ~1k employees and minimum estimated time savings of $43k, we calculate

How? The Technical Component

NLP: Natural Language Processing

• Natural Language Processing is the scientific study of language by a computer or AI.
• It is the method by which large amounts of text are consumed and analyzed for the purposes of identifying predictive patterns that facilitate the understanding and effect of the language.
• This NLP was developed by integrating university linguistic studies based on text learning analytics, text mining and AI.
• Algorithms build a model based on sample data, known as training data, in order to make predictions or decisions without being explicitly programmed to do so.

For an opportunity to see Textmetrics in action