Kimola Glossary

This glossary defines the terms you see in Kimola, the AI platform that turns customer feedback into enterprise-grade market research. In Kimola, you analyze customer feedback in a Report or a Space. Models classify every comment into themes and sentiment, Dimensions ask it research questions, and the findings become Signals and Action Plans. Terms are listed from A to Z. Each one has a short definition and a link to the article that explains it in detail. Terms for features that are available only in the Enterprise package say so in their definition.

A

Action Plan

An Action Plan is a set of up to 12 actions that Kimola writes from a Report for a Profile, such as your company, a brand or a client. Kimola forms hypotheses from the Report, tests each one against the customer feedback and drops any action the evidence doesn't support. The remaining actions are scored and ranked. You find Action Plans under Actions in a Report. See What is an Action Plan?

Airset

An Airset is a dataset of customer feedback that you collect from one of 30+ supported platforms with Kimola's browser extension. You can view and download your Airsets on the Airsets page, or create a Report from any of them. See What is an Airset?

Alert

An Alert is a rule in a Feed that emails your team when something important happens. An Alert can fire when mentions contain keywords you choose (Contains keywords), when the total number of mentions reaches a number (Total Volume Increase), or when a social media user with more than a set number of followers mentions you (Influencer Involvement). Available in paid packages, not in the free Starter package. See What is an Alert?

B

Batch

A batch is one dataset of customer feedback uploaded to a Space. Each Space grows batch by batch, and every batch is classified with the same Funnel. In the app, batches are listed on the Space's Files page and called files. See Add Files to a Space. Available in the Enterprise package.

C

Category

A category groups related labels of a model under one name, such as Claims. Kimola can draft categories from your labels, and you can rename them, describe them and move labels between them. In Reports and Spaces, labels appear under their categories, but feedback is still classified with the labels themselves. Available in paid packages, not in the free Starter package. See Organize Labels into Categories.

Classifier

A classifier is a model that assigns labels to every piece of customer feedback. A content classifier, such as the Automatic Classifier, finds what customers talk about. A sentiment classifier, such as the Sentiment Classifier, finds how they feel about it. A classifier can be a pretrained model or a custom model. See What is a Model?

Custom Model

A custom model is a model you create with your own labels, from a prompt, a training dataset or an existing Report. A prompt-based custom model can also work without predefined labels: right before it classifies a dataset, it answers your question to discover the labels from that data. Custom models give your organization a shared language for insights. Available in paid packages, not in the free Starter package. See What is a Custom Model?

Customer

A Customer is a person or organization that Kimola recognizes by the value in a Space's Customer Column, such as a customer code, an ID, a number or an email address. The same value means the same Customer across all your Spaces, so you can see everything one customer has told you in one place. Available in the Enterprise package. See What is a Customer?

Customer Column

The Customer Column is the column in a Space's files that tells Kimola who wrote each piece of feedback. You can choose it when you upload the first file, or later on the Space's Files page. When you choose it later, existing records get their Customers without being classified again. Available in the Enterprise package. See Add a Customer Column to a Space.

D

Dataset

A dataset is a table of customer feedback that you analyze in Kimola, such as an .xls, .xlsx, .csv or .tsv file you upload. You choose the column that holds the feedback text, and you can also choose a URL column and a Date column. Other columns, such as country or product, let you filter and compare the results. See Create a Report by Uploading a Custom Dataset.

Dimension

A Dimension is a research question that Kimola asks of every piece of customer feedback. For example, the Pain Points Dimension asks "What goes wrong or creates difficulty?" and the Motivations Dimension asks "Why is this chosen or preferred?" You can add Dimensions to a Report when you create it, or to a Funnel for your Spaces. See What is a Dimension?

Dynamic Classification

Dynamic Classification is how Kimola finds themes in customer feedback without a predefined list of categories. Kimola processes the feedback before and after grouping it, so each theme comes from the data and gets a clear name that describes what customers say. The Automatic Classifier works this way. See How Kimola Uses AI for Research.

F

Feed

A Feed keeps collecting public mentions and reviews from social media, review sites and the web. It tracks the Keywords and Links you give it, and you can filter its mentions, export them or create a Report from them. Available in paid packages, not in the free Starter package. See What is a Feed?

File

A file is one dataset of customer feedback added to a Space, also called a batch. The first file sets the Space's schema: the Date column, the text fields to classify, the other columns to filter and compare by, and an optional Customer Column. Every file is classified with the Space's Funnel, and a file can be taken back together with its records. Available in the Enterprise package. See Add Files to a Space.

Funnel

A Funnel is the set of custom models that a Space classifies feedback with: a content classifier, a sentiment classifier and Dimensions, plus the language that Dimension analyses, Signals and Action Plans are written in. Because every file uses the same Funnel, results stay comparable over time and across Spaces. Funnels are shared by the whole team. Available in the Enterprise package. See What is a Funnel?

I

Integration

An integration connects Kimola to another tool. Zendesk, Intercom and Google Sheets bring customer feedback into a Report, Facebook pages and Radaar bring data into a Feed, and Jira receives Tasks. You manage integrations under Account › Integrations. Available in paid packages, not in the free Starter package; the Jira integration is available in the Enterprise package. See What are Integrations?

K

Keyword

A Keyword is a word, phrase or hashtag that a Feed tracks across the web, to capture what people say about a topic such as your brand, a product or a competitor. Available in paid packages, not in the free Starter package. See What is a Keyword?

Kimola On-Premises

Kimola On-Premises is the version of Kimola that runs inside your organization's own infrastructure, under a license. Some features differ from the cloud version at app.kimola.com. For example, Kimola On-Premises has no Feeds and doesn't offer the third-party data integrations. See What is Kimola On-Premises?

L

Label

A label is one value that a classifier can assign to a piece of feedback, such as a theme or a sentiment. When you read results, labels appear as themes. When you build or edit a model, you work with its labels, and a custom model's labels can be grouped into categories. See Manage Labels in Reports.

Link

A Link is the address of a web page that Kimola collects customer feedback from. When you create a Report, you paste a link to a product, business or app page, such as an Amazon product, a Google Business listing or an App Store app, and Kimola analyzes the feedback on it. See Create a Report by Adding Links. In a Feed, a Link is a page that the Feed keeps tracking for new reviews, comments or posts, instead of a Keyword. Links in Feeds are available in paid packages, not in the free Starter package. See What is a Link?

M

Mailing

A Mailing is a scheduled email that sends a Feed's latest mentions to you and your colleagues at an interval you choose. It can cover all channels or only social media, reviews or the web, include only featured mentions, and attach an Excel file with all the conversations. Available in paid packages, not in the free Starter package. See What is a Mailing?

Multi-label Classification

Multi-label classification gives one piece of feedback every theme it mentions, and each theme gets its own sentiment. With single-label classification, each piece of feedback gets one label. You choose between them when you create a Report or a Funnel. See Multi-label vs. Single-label Classification.

N

NPS

In a Space, NPS is the share of positive feedback minus the share of negative feedback, from −100 to 100, based on the Funnel's sentiment classifier. It is calculated from what customers write, not from 0–10 survey ratings, so it is not the same as the Net Promoter Score of a survey. Available in the Enterprise package. See How NPS is Calculated in Spaces.

P

Pretrained Model

A pretrained model is one of Kimola's ready-made models, which you can use without any setup. They include the Automatic Classifier, the Sentiment Classifier and classifiers built for industries such as e-commerce, mobile apps, software and banking. See What is a Pretrained Model?

Profile

A Profile is the organization an Action Plan is written for: your company, one of your brands or a client you work for. A few words on what it does, what it sells and to whom make the actions more specific. Profiles are shared with your team and managed under Account › Profiles. See Manage Profiles.

Project

A Project groups related Feeds, Reports and Models in one place to keep your work organized. Deleting a Project doesn't delete the items in it. See What is a Project?

Q

Query

A query is Kimola's unit of usage. Each analyzed row of up to 500 characters counts as one query. To see how many queries each package includes, visit the pricing page. See How to Calculate Queries.

R

Report

A Report is a finished analysis of one dataset of customer feedback: a file you upload, links, an integration, an Airset or a Feed. You choose its models and Dimensions. Kimola structures every comment into themes, sentiment and Dimensions, looks for Signals, and lets you write Action Plans from the results. See What is a Report?

Restriction

A Restriction sets how many mentions a Feed can collect from one host domain. It keeps a single site from filling your Feed and helps you stay within your package limits. Available in paid packages, not in the free Starter package. See What is a Restriction?

S

Satisfaction Score

The Satisfaction Score and the Dissatisfaction Score show how positive or negative customers are about a theme. Unlike a plain percentage, they take the number of comments into account along with the share, so small and large themes can be compared fairly. Kimola uses them to rank your best and worst performing themes. See Performance in Reports.

Sentiment

Sentiment is how positive or negative a customer is in their feedback. Kimola finds the sentiment of each piece of feedback and, with multi-label classification, of each theme in it, so one comment can be positive about one theme and negative about another. This is called aspect-based sentiment. See How Kimola Uses AI for Research.

Signal

A Signal is a statistically tested finding in a Report or a Space. Differences show where a theme or Dimension is mentioned more, or more negatively, in one group than in the rest, such as one country or one product. Trends show how mentions changed in recent periods. Signals are scored and ranked, and you can mark each one as Known, New, Unimportant or Artifact. See What is a Signal?

Space

A Space is a channel of customer feedback that keeps growing, such as a call center, an NPS survey or social media. You add a new file whenever new feedback arrives, and every file is classified with the same Funnel, so results stay comparable over time and across Spaces. NPS, Customers and Tasks build on Spaces. Available in the Enterprise package. See What is a Space?

T

Table

A Table is a cross-tab view of the feedback in a Report or a Space. You build it by dragging columns, classifiers and Dimensions onto it. A column can group records by its values or, if it holds numbers, be measured. You can save a Table and open it again later. See Tables in Reports.

Task

A Task turns a finding into work for one teammate, with the feedback and numbers behind it kept as evidence. You can create a Task from a Signal, a piece of feedback, a filtered view, a theme, a Customer or a Table, or write your own. After a Task is done, Kimola reads the same feedback again two and four weeks later to see whether customers noticed. Tasks can also be sent to Jira. Available in the Enterprise package. See What is a Task?

Theme

A theme is a topic customers talk about in their feedback, such as delivery time or app performance. Themes are the labels a content classifier assigns. The Automatic Classifier finds them in the data itself, while a custom model uses the themes you define. See How Kimola Uses AI for Research.