Skip to the content
Documentation
Help Start
Reading in:
  • German: 9,090 of 9,090 pages
About the help in other languages
All help pages

Running content modeling workshops

How to run workshops that turn your content into content classes, attributes, a content tree, sections, states, languages and roles; agendas, exercises and a class sheet to copy.

Download as PDF

Address: https://doc.exponential.earth/exphelp/viewer/help/leadership/content-modeling-workshops.md

On this page

Running content modeling workshops

This page is for project managers who facilitate content modeling, and for the editors and developers who take part. It answers: how to prepare and run the workshops, which questions to ask, how each answer maps to a concept of Exponential (class, attribute, datatype, node, section, state, language, role, layout), and what the workshops must produce so that developers can build and editors can recognize their content. The product terms come from The content model and editing content.

Contents

  1. Why content modeling matters in Exponential
  2. The concepts in one table
  3. Who takes part
  4. Preparation
  5. Workshop 1: content inventory and types
  6. Workshop 2: attributes and datatypes
  7. Workshop 3: the tree, relations and taxonomy
  8. Workshop 4: permissions, stages and languages
  9. Workshop 5: pages and layouts
  10. The class sheet to copy
  11. After the workshops
  12. Common pitfalls

Why content modeling matters in Exponential

"As a CMS its most notable feature is its fully customizable and extendable content model" (README, quoted on Who is the Exponential Foundation?). In Exponential you define your own types of content instead of fitting your content into fixed pages and posts. That freedom is the reason to model carefully: every class you agree on becomes forms that editors fill in every day, templates that developers write, and permissions that administrators assign.

The concepts in one table

ConceptWhat it isDecided inRead
Content classA type of content: a name, an identifier and a list of attributes (for example Article: Title, Intro, Body, Image)Workshop 1Content model, section 1
Class groupA folder of classes in the class editor (Content, Users, Media, Setup)Workshop 1Class groups
Attribute and datatypeOne field of a class and its kind: one line, several lines, rich text, image, relation ...Workshop 2Content model, section 1
Content objectOne piece of content, an instance of a class, with versions and languagessame
Node, locationA place of an object in the content tree; one object can have severalWorkshop 3same
SectionA group of objects for permissions, navigation and designs; not a place in the treeWorkshop 4Sections
Object stateHow far along content is (review, legal clearance, embargo, featured)Workshop 4Object states
WorkflowSteps such as approval that run before or after an operationWorkshop 4Workflows
Content languageA language content is written in; shown per siteaccessWorkshop 4Content languages
Role and policyWho may do what, limited by section, class, subtree, language, stateWorkshop 4Roles and policies
Layout, zone, block, ruleHow a page is arranged, and which pages get which layoutWorkshop 5The layouts editor

Who takes part

PersonWhyWorkshops
Project managerFacilitates, keeps the decision log, guards the scopeall
Lead editorKnows the content and the daily workall
One or two editorsWill fill in the forms; spot what is missing1, 2, 5
Lead developerKnows what each decision costs to build; maps answers to classesall
DesignerKnows which fields the design needs2, 5
AdministratorRoles, sections, languages, siteaccesses4
Translator or language ownerFor multilingual sites4

Keep the group small enough to decide. Invite the others to a review of the results instead.

Preparation

  • To do: A demonstration installation everyone can open, with a site package that has demo content (sevenx_multisite is the default of exp:install; 4.3 Site packages). Showing real classes in Setup > Classes makes every concept concrete.
  • To do: A content inventory: a list of the content that exists today (pages, documents, products, people, events), with rough counts and an owner per area.
  • To do: Ten real examples of content, printed or on screen, of different kinds.
  • To do: The requirement list from discovery (Planning an Exponential project).
  • To do: Copies of the class sheet.
  • To do: A decision log (Project templates).

Workshop 1: content inventory and types

Goal. A list of content classes with a one-sentence purpose each.

Agenda:

  1. Show the product (15 minutes). The developer opens a class in Setup > Classes and creates an object of it, following Content model, sections 2 and 3. Everyone sees that a class is a form and an object is a filled-in form.
  2. Sort the examples (30 minutes). Lay out the ten examples. Group those that have the same fields. Each group is a candidate class.
  3. Name the classes (20 minutes). A name people read and an identifier code uses (article, event). Note the object name pattern: which field names an object in lists (for example <title>).
  4. Check against the inventory (20 minutes). Does every item of the inventory fit one class? Items that fit none are a new class or a sign that two classes should be one.
  5. Reuse (15 minutes). Which classes already exist in the site package? Reusing a shipped class saves template and translation work.

Output: the class list, each class with name, identifier, purpose, owner, estimated number of objects.

Tip

Ask "what would an editor search for?" rather than "what does the page look like?". A class describes content, not a page layout; the layout is decided in workshop 5.

Workshop 2: attributes and datatypes

Goal. For each class, its attributes with their datatype, whether required, and whether translated.

The datatypes available for new attributes are the AvailableDataTypes[] list of settings/content.ini, 35 entries in the shipped list (Content model, section 1); extensions add more. Map the editors' words to them:

Editors sayDatatype identifier (settings/content.ini)
a title, a name, one line of textezstring
several lines of plain texteztext
formatted text with headings, links, imagesezxmltext (rich text, edited in the online editor)
a date; a date and time; a timeezdate, ezdatetime, eztime
yes or noezboolean
a numberezinteger, ezfloat
one choice from a listezselection
a link to another item; severalezobjectrelation, ezobjectrelationlist
a pictureezimage
a file to download; audio or videoezbinaryfile, ezmedia
a web address; an e-mail addressezurl, ezemail
keywordsezkeyword
a table of valuesezmatrix
a priceezprice, ezmultiprice
a countryezcountry
a user accountezuser

Extensions add others, for example tags (eztags), map locations (ezgmaplocation), star ratings (ezstarrating) and spam protection for forms (recaptcha, hcaptcha) (Extensions, themes and packages).

For each attribute decide:

  • To do: Name and identifier.
  • To do: Datatype.
  • To do: Required? Image fields can require alternative text (Content model, section 9); decide it here, it is an accessibility requirement.
  • To do: Translated? For a multilingual site: is the value the same in every language (a date, a price) or written per language?
  • To do: Searchable? Will people search for it?
  • To do: Shown where? Full view, list view, both, nowhere (internal).
  • To do: Example value from the real examples.

Output: one completed class sheet per class.

Workshop 3: the tree, relations and taxonomy

Goal. Where content lives, how items point to each other, and how content is classified.

Questions:

  1. The tree. Draw the top two or three levels of the content tree. Which class may be created below which? Content that belongs in several places gets several locations (one object, several nodes), not copies.
  2. Relations. Which items refer to which (an article to its author, an event to its venue)? Each is an object relation attribute. Decide what happens when the target is removed; the content checks list "broken relations" afterwards (Content checks).
  3. Taxonomy. Is classification free (keywords) or controlled (tags managed by someone)? Controlled tags are eztags (eztags); check the database first, because eztags is not usable on MongoDB (MongoDB limits).
  4. Media. Where do images and files live: in the media library, below the content that uses them, or both?
  5. Addresses. Which URL should an item have? URL aliases follow the tree; global aliases and wildcards exist for exceptions (URL aliases).

Output: a tree diagram (on paper is fine), a relation list, a taxonomy decision.

Workshop 4: permissions, stages and languages

Goal. Sections, states, workflows, languages and the role plan.

QuestionProduct answerRead
Are there groups of content only some people may read or edit (members area, intranet department)?A section per group; roles limited by sectionSections
A new installation has five sections (Standard, Users, Media, Setup, Design); which do we add?New sections; note that the Anonymous role reads only Standard by defaultSections, section 2
Does content pass stages, with different people allowed in each stage?A state group with states; policies limited by stateObject states
Must someone approve before publishing?A workflow with an Approve event and a trigger; the workflow cronjob must runWorkflows
Which languages, and which is the main language of each item?Content languages; SiteLanguageList per siteaccess; "always available" itemsContent languages
Who may do what?Roles assigned to user groupsRoles and policies

Note

Sections, states and the tree are three independent ways to group content. "Sections say where content belongs ..., and the content tree says what it sits under. States say how far along content is" (Object states). Use the one that matches the question, not all three for everything.

Output: section list, state groups, workflow list, language list, role plan (role, groups, policies, limitations).

Workshop 5: pages and layouts

Goal. Which pages editors arrange themselves, and which are fixed templates.

With Exponential Layouts, "a layout is a saved arrangement of blocks in zones, and a rule decides which pages get it" (Who is the Exponential Foundation?). In the workshop:

  1. Open the Layouts tab of the demonstration installation and walk "In action: the layouts editor" of the admin tour.
  2. List the page types (home, landing pages, section front pages, article pages).
  3. For each: fixed template, or a layout editors change? Which zones? Which blocks?
  4. Which blocks are the same on many pages (header, footer, a prefooter)? They belong in a shared layout, linked from the others (The layouts editor).
  5. Who may change layouts? The Layouts tab is shown only to roles that may work with layouts.

Output: page type list with layout decision, zone and block list, shared layout list.

The class sheet to copy

Class: <Name>                         identifier: <identifier>        group: <Content | Media | ...>
Purpose: <one sentence>
Owner (editor): <name>                Expected number of objects: <n>
Created below: <parent classes>       May contain: <child classes>
Object name pattern: <<attribute>>    Section: <section>               States: <state group or none>
Approval before publishing: <yes/no, by whom>

| # | Attribute name | Identifier | Datatype       | Required | Translated | Searchable | Shown in        | Example value |
|---|----------------|------------|----------------|----------|------------|------------|-----------------|---------------|
| 1 | Title          | title      | ezstring       | yes      | yes        | yes        | full, list      | <example>     |
| 2 | Intro          | intro      | ezxmltext      | no       | yes        | yes        | full, list      | <example>     |
| 3 | Image          | image      | ezimage        | no       | no         | no         | full            | <file>        |
| 4 | <name>         | <id>       | <datatype>     | <y/n>    | <y/n>      | <y/n>      | <where>         | <example>     |

Open questions: <list>
Decided on: <date>   by: <names>

After the workshops

  1. Write it up in one document: class sheets, tree, relations, taxonomy, sections, states, workflows, languages, roles, page types. Version it in the project repository next to the code.
  2. Review with editors who were not in the room. Ask each to place five real items in the model.
  3. Prototype: a developer creates two or three classes on the test installation (a class is created in minutes, Content model, section 2). Editors enter real content into them.
  4. Adjust and freeze the model for the build. Later changes go through the change process of Estimating and scoping.
  5. Plan migration and entry from the class sheets (Planning, phase 4).

Common pitfalls

PitfallWhy it hurtsRemedy
One class per page designMany classes with the same fields; templates multiplyModel content; arrange pages with layouts or view templates
One "generic page" class for everythingNothing can be listed, filtered or reusedSplit by what editors search for
Copies instead of locationsTwo objects drift apartOne object with several locations
Sections used as foldersSections are not places in the treeUse the tree for structure, sections for permissions
Languages decided lateTranslation flags, siteaccesses and roles changeDecide languages in workshop 4
Too many languages added "just in case"Each language is a bit of a limited mask (at most 62) and removing one is only possible when nothing uses itAdd a language when it is needed (Content languages)
Alternative text left optionalAccessibility gaps found in QAMake it required in workshop 2

Become an Exponential Partner

Partners share their information, provide funding, contribute back and develop the products with us, and that is how Exponential grows. A partner website and dashboard are planned for mid-2027: please become a partner today to speed up our growth rate and land new features faster.

Something wrong on this page? Report an error

Documentation license

This documentation is free documentation, licensed under the GNU Free Documentation License 1.3 or later.

Exponential Copyright © 1998-2026 Exponential Foundation and 7x