Product & Customers
Explain Natter in 1-2 sentences
Varies, something along the lines of: Natter is a conversational AI platform that delivers leaders the most human insights possible
What devices can people access Natter on?
Natter is accessible through either a desktop/laptop computer or, if the employee is a frontline worker, on a smart phone capable of facilitating a video-call based experience (e.g. Google Meet / Zoom / WhatsApp Video Call / Facebook Messenger Video Call).
How does integrating Natter compare with other HR software?
Super simple! We have a very simple onboarding process that takes 10 minutes by the time we have relevant company info.
Where can I find more information about Natter's information security and data protection posture?
There are two public-facing resources intended to build confidence with prospects and existing customers:
i) Natter Security Webpage: This is very basic, but could also be shared alongside the Whitepaper below: https://www.trynatter.com/security-privacy
ii) Natter Whitepaper, titled "Our Commitment to Information Security & Data Protection": Which is intended as a sales and marketing asset to build confidence/
What data is stored by your AI Engine after analysis takes place?
Our AI Engine does not retain any information after invocation. The inputs are sent to the system, and outputs retrieved and stored as needed in our standard backend database.
Name 3 of Natter's functionalities
i) allows tens of thousands of people to connect, ideate and respond to key questions, all at once
ii) captures and analyzes millions of your peoples' ideas and responses through real-time video conversations and AI
iii) maximises participation and diversity of perspectives in the world's most scalable off the shelf focus group
If English is not someone's first language, can they use Natter?
Typically, all of your employees that commonly join video calls and meetings in business English will be able to join and participate fully in a Natter event. Where you have a workforce population that speaks one language (e.g. Brazil / Portugese), we are able to add the survey questions and conversation topics in whichever language would be spoken by participants, in place of English.
What is needed from the customer to onboard them?
Natter has been designed as a light touch system that can be accessed by anyone with an internet connection and a web browser. No downloads, installations or integrations are required. Simply set up a Community on Natter, where we require:
i) Company name (e.g. "Legal & General")
ii) Company corporate branding HEX code (e.g. #0000FF)
iii) Community Tags (e.g. tags should fall under 4 categories where the default are: Location, Function, Tenure, Seniority)
Where is customer data hosted?
All customer data is hosted in the cloud through Amazon Web Services (AWS) in the UK at hosting location eu-west-2. Availability Zones (AZs) are configured to make use of a single data center more highly available, fault tolerant, and scalable.
What technology / AI do you use to synthesise and make sense of free text responses?
The topic detection and classification of free text into dominant themes is performed by OpenAI Generative Pre-trained Transformer 4 (GPT-4). GPT-4 is a multimodal large language model created by OpenAI, and the fourth in its series of GPT foundation models.
Name 3 common use cases for Natter
Town Halls / All Hands, Employee Survey Deep Dives, Learning & Development, DE&I, ERGs, Health & Wellbeing, Innovation, Company Culture, Change Management, Employee Onboarding, OKR Planning & Review, Benefits & Compensation
Why does Natter only have 1:1 functionality and not 3/4/5 person groups?
Our initial user testing of the platform and product concepts validated the importance of being able to facilitate a 1:1 conversation dynamic at scale. No other format is able to provide:
i) A safe space: Privacy is easier to maintain in a one-on-one conversation. Participants can discuss sensitive or personal topics
ii) More meaningful connection: A deeper emotional connection and trust is built between the participants, encouraging openness and vulnerability
iii) Enhanced empathy: In a more intimate setting, participants can more easily pick up on non-verbal cues such as body language and tone of voice
iv) Faster decision making: Without the need to achieve a group consensus, the two individuals can quickly discuss options, weigh pros and cons
v) Reduced peer pressure: In a one-on-one setting, individuals are less likely to be influenced by group dynamics and peer pressure
All these attributes - only possible through intimate, 1:1 conversations as the format - lead to a greater and more considered quantum of insightful data capture.
What are community tags and what is their purpose?
Community tags are chosen by the customer and fed into the Natter platform by Natter Support during Community creation. Each end user must select 1 community tag within each category to set up their account. They have a dual purpose on Natter, being:
i) Anonymous Insights Benchmarking: Tagging responses allows customers to understand diverse perspectives whilst maintaining anonymity
ii) Powers our Matching Algorithm: Natter seeks to facilitate conversations that would not normally take place. Our matching algorithm seeks to drive perspectives sharing between people outside of for example, their function or geography (e.g. a senior finance leader in Mumbai with a junior sales rep in Miami)
How do we manage around VPNs and Firewalls?
We recommend whitelisting. Customers may be put off knowing we suggest they whitelist, to quell their concerns reference:
i) Only 5% of our customers whitelist, typically financial services companies or consultancies that operate in the defence or governmental sectors
ii) We have completed whitelisting for large, complex enterprise customers in as little as 10 working days
iii) Should you wish to proactively whitelist, our technical team to discuss, answer questions and support you (i.e. James Stevens)
What employee data has been used for model training?
Our customers' data has not and will not be used for model training. As is stated in OpenAI's documentation: "we don’t use content from our business offerings such as ChatGPT Team, ChatGPT Enterprise, and our API Platform to train our models"
Beyond the offerings of OpenAI, Natter has not trained and does not use any supplementary models.
Name 5 active Natter customers
Deloitte, Miro, Legal & General, Marqeta, AlphaSense, AMS, Prolink, Agilent, Miro, Rakuten, Anglo, Marqeta, SAP, Doorfeed
Can Natter be used for non-employees (e.g. customers of our product)?
Yes. Natter can be used to capture insights and conversational intelligence from anyone with an internet connection and a laptop/desktop or mobile device, including:
i) Consumers: Engage everyday consumers to gather feedback on any commercial strategy of your choice (e.g. product launch/feedback, marketing campaign)
ii) Voters: Gather insights on how to engage swing voters or mobilise your political base (e.g. policy formulation/feedback)
iii) Members: Engage members of trade associations, members clubs or industry bodies
How many conversations should you have at a Natter Event?
Most Natter Events last between 45 and 60 minutes (or sometimes up to 90 minutes). Given the time needed for any first time participants to set up their accounts and join, alongside the need to screen share content or provide other guidance, we recommend:
i) 45 minute event: 2 x 10 minute conversations
ii) 60 minute event: 3 x 10 minute conversations
iii) 90 minute event: 4 x 10 minute conversations
We recommend each conversation lasts 10 minutes, but Natter can time conversations on settings: 2, 3, 5, 7, 10, 15, 20 and 30 minute settings.
What security accreditations does Natter have?
At Natter, security is an intrinsic part of our DNA. We believe security is everyone's responsibility and safeguard customer data as if it were our own. Our compliance is monitored by a range of enterprise-grade monitoring tools, including Drata, that allows us to continuously monitor compliance. Natter is accredited to (or compliant with) the following:
i) ISO 27001: Natter is accredited to ISO 27001, the most widely recognised information security standard globally, where our compliance is audited by a third party auditor. Our certificate of certification can be found: https://natter.box.com/s/ng617mg1sx9h974pi99ht08vql0dolex
ii) UK + EU GDPR: Natter is UK-headquartered and compliant with both the UK and EU GDPR data protection regulations (and their rules on data protection, privacy and transfer)
How can Natter analyse sentiment?
Natter can capture sentiment in two ways:
i) Survey Questions: Natter has the functionality to capture sentiment through Survey Questions, that can be posed at any stage of a Natter event. Capture sentiment through use of a range of Question Types (Single Choice, Multiple Choice, Slider, Ranking and NPS) and a range of linear scales to assess sentiment on any subject and any form that matters to you.
Conversations: Natter captures considered feedback from participants through targeted, meaningful conversations. The value of these insights goes above and beyond basic sentiment analysis to generate actionable ideas to address such sentiment
What three customers have we interviewed for case studies?
Miro
Marqeta
AMS
Are there any regional restrictions or considerations we should be aware of?
Natter is built with inclusivity and accessibility front of mind and anyone with an internet connection and a laptop/desktop or smart phone can participate.
NOTE: Like similar Western technologies (e.g. Zoom, Facebook), Natter may struggle to be used reliably and predictably within Mainland China. However unlike Zoom, Natter is not blocked in China. This is not a message you should proactively deliver to a prospective customer unless they directly seek to address it.
What is a DPA and how do we address a customer asking for one?
Having a Data Processing Agreement (DPA) in place is a requirement to be GDPR compliant and governs how data is transferred between different legal entities (e.g. between Natter and a customer). Any customer who has signed an Order Services Form (i.e. commercial agreement / sales contract) with Natter will benefit from a number of clauses within the T&Cs of that contract that define how data is processed. Hence, from a basic legal perspective, they are covered. However, many of the customers we work with are complex and risk averse and request we sign an additional DPA before going live with them. If this is the case, request the template DPA they would like us to sign and James Stevens can fill in the requirements.
Note: signing enhanced DPAs slows the sales cycle and delays customer onboarding and is only really relevant for customers in the UK and EU (GDPR is not a regulation that covers the USA). Therefore, unless the customer requests we sign their DPA, we should assume they are comfortable with the data processing clauses in place in the Natter Services Order Form.
How do you guard against or mitigate bias?
The topic detection and classification of free text into dominant themes typically undergoes a human-in-the-loop audit: a "mechanical turk" to ensure a higher rate of accuracy and reliability. This allows for any bias identified during this phase to be corrected. Bias mitigation is an ongoing area of research for our AI Engine service provider and where we can, we provide feedback that may support them in continuing to improve.