Skip to content
Language اللغة
Intermediate workshop · 2 hours · Online ورشة بمستوى متوسط · ساعتان · عن بُعد

Beyond chatbots — AI for media and marketing. ما بعد بوت الدردشة — الذكاء الاصطناعي للإعلام والتسويق.

An intermediate, two-hour session on selected AI use cases in media and marketing. The goal is to widen your awareness of what's possible beyond prompting a chatbot — sourced research, structured analysis, synthesis across many inputs at once, and outputs you can actually act on. جلسة بمستوى متوسط لمدة ساعتين تتناول حالات استخدام مختارة للذكاء الاصطناعي في الإعلام والتسويق. الهدف هو توسيع وعيك بما هو ممكن بعد مرحلة الكتابة في بوت الدردشة — بحث موثّق المصادر، تحليل منظّم، توليف عبر مدخلات متعددة في وقت واحد، ومخرجات يمكنك التصرف بناءً عليها فعلاً.


2h
A focused, two-hour live session جلسة مباشرة مركّزة لمدة ساعتين
4
Core topics, theory plus practice أربعة محاور أساسية: نظرية وتطبيق
EN ع
Bilingual content — toggle any time محتوى ثنائي اللغة — يمكنك التبديل في أي وقت
100+
Marketing & media professionals attending أكثر من 100 محترف في التسويق والإعلام
Agenda جدول الأعمال

Two hours, deliberately paced. ساعتان بإيقاع مدروس.

Each block pairs the marketing science with a concrete AI exercise. Where running a tool live would burn time, the prompt and result are pre-prepared so the flow stays clean. يجمع كل قسم بين علم التسويق وتمرين عملي على الذكاء الاصطناعي. حين يكون تشغيل الأداة مباشرةً مكلفاً للوقت، تكون الموجّهات والنتائج محضّرة مسبقاً لإبقاء التدفق سلساً.

## Blockالمحور Timeالوقت On this pageفي هذه الصفحة
0 Welcome, objectives & agendaالترحيب، الأهداف وجدول الأعمال 10 min10 دقائق Live onlyفي الجلسة فقط
1 On AI — foundation conceptsعن الذكاء الاصطناعي — مفاهيم تأسيسية 15 min15 دقيقة Live onlyفي الجلسة فقط
2 Media & marketing use cases — overviewحالات استخدام الإعلام والتسويق — نظرة عامة 5 min5 دقائق Live onlyفي الجلسة فقط
I Analyzing & creating contentتحليل المحتوى وإنشاؤه 35 min35 دقيقة 10 units + briefing10 وحدات + إحاطة
II Customer experience journey mapping (CXJM)رسم خرائط رحلة تجربة العميل (CXJM) 20 min20 دقيقة 15 units15 وحدة
III Beyond chatbots — broader AI capabilitiesما بعد بوت الدردشة — قدرات الذكاء الاصطناعي الأوسع 20 min20 دقيقة Live onlyفي الجلسة فقط
4 Closing & Q&Aالختام والأسئلة 15 min15 دقيقة Live onlyفي الجلسة فقط
Totalالمجموع 120 min120 دقيقة
Workshop block I · 35 minutes المحور الأول من الورشة · 35 دقيقة

Analyzing & creating content. تحليل المحتوى وإنشاؤه.

Three topics. We start with NotebookLM to gather the sources we'll work from. Then we use Wheeler's framework as the lens for analysis, and finally build the message house that turns that analysis into consistent output. ثلاثة مواضيع. نبدأ بـ NotebookLM لجمع المصادر التي سنعمل عليها. ثم نستخدم إطار Wheeler كعدسة للتحليل، وأخيراً نبني بيت الرسالة الذي يحوّل ذلك التحليل إلى مخرجات متّسقة.

Block I · Topic A · NotebookLM المحور I · الموضوع أ · NotebookLM

From scattered sources to a focused brief. من مصادر متفرّقة إلى ملخّص بحثي مركّز.

Topic A · Briefing الموضوع أ · إحاطة Briefing pack pending حزمة الإحاطة قيد الإعداد

What you'll see in the live session ما ستراه في الجلسة المباشرة

We open Block I with a hands-on tour of NotebookLM. You provide a handful of sources — articles, PDFs, transcripts — on the topic we'll work with through the rest of the workshop, and NotebookLM turns them into something you can use in minutes. نفتتح المحور الأول بجولة عملية في NotebookLM. تقدّم حفنة من المصادر — مقالات، ملفات PDF، نصوص — حول الموضوع الذي سنعمل عليه طوال الورشة، ويحوّلها NotebookLM إلى شيء يمكنك استخدامه في دقائق.

What we cover ما سنغطّيه

  • Adding sources to a notebook (web pages, PDFs, Google Drive).إضافة المصادر إلى دفتر ملاحظات (صفحات ويب، PDF، Google Drive).
  • Generating an audio overview — a two-host podcast made from your sources.إنشاء ملخّص صوتي — بودكاست بمضيفَين مأخوذ من مصادرك.
  • Turning sources into a structured study guide and an infographic-style brief.تحويل المصادر إلى دليل دراسة منظّم وإحاطة على شكل إنفوغرافيك.
  • Asking grounded questions — answers cite the exact passages from your sources.طرح أسئلة مرتكزة — إجابات تستشهد بالمقاطع الدقيقة من مصادرك.
  • Hand-off: feeding the NotebookLM output into Claude or Gemini for further analysis.التسليم: تمرير مخرجات NotebookLM إلى Claude أو Gemini لمزيد من التحليل.
Why it sits here in the flow: the rest of Block I assumes we have a clear, sourced view of the topic. NotebookLM gets us there fast — without burning live-session time on manual reading. لماذا يقع هنا في التدفق: بقية المحور الأول تفترض أن لدينا رؤية واضحة وموثّقة المصادر للموضوع. NotebookLM يوصلنا إلى تلك النقطة بسرعة — دون استهلاك وقت الجلسة المباشرة في القراءة اليدوية.

Resources for this unit مصادر هذه الوحدة

Open NotebookLM in one tab, drop the two source documents below into a new notebook, and you're ready to follow along. افتح NotebookLM في علامة تبويب، أضف وثيقتَي المصدر أدناه إلى دفتر ملاحظات جديد، وستكون جاهزاً للمتابعة.

Outputs from this NotebookLM مخرجات هذا الـ NotebookLM

After loading the two source PDFs above into a notebook, NotebookLM produced an Arabic audio overview and two infographic briefs — one in each language. They are embedded below. بعد تحميل ملفّي PDF أعلاه إلى دفتر الملاحظات، أنتج NotebookLM ملخّصاً صوتياً بالعربية وإحاطتَين على شكل إنفوغرافيك — واحدة بكل لغة. كلها مُضمَّنة أدناه.

Audio Audio overview — Arabic ملخّص صوتي — بالعربية

Two-host conversational summary generated by NotebookLM directly from the source PDFs. حوار بين مضيفَين كملخّص أنتجه NotebookLM مباشرةً من ملفات المصادر.

English infographic — NotebookLM output
Image Infographic — English إنفوغرافيك — بالإنجليزية

One-page visual brief generated by NotebookLM. Click the image to open full size. إحاطة بصرية من صفحة واحدة أنتجها NotebookLM. انقر على الصورة لفتحها بالحجم الكامل.

Arabic infographic — NotebookLM output
Image Infographic — Arabic إنفوغرافيك — بالعربية

Note: the Arabic infographic contains spelling mistakes — kept as-is to show that AI outputs still need a human pass. ملاحظة: الإنفوغرافيك العربي يحتوي على أخطاء إملائية — أُبقيت كما هي لإظهار أن مخرجات الذكاء الاصطناعي لا تزال تحتاج إلى مراجعة بشرية.

Walk-through جولة عملية

Researching a topic outside your expertise — AI & copyright law البحث في موضوع خارج اختصاصك — الذكاء الاصطناعي وقوانين حقوق الملكية الفكرية

A practical demo of a different NotebookLM workflow: when you need to come up to speed on a domain you don't know. Here, the goal is to understand legislative trends in IP rights as they relate to AI — across many countries, multiple legal traditions, and several stakeholder perspectives. عرض عملي لسير عمل مختلف على NotebookLM: حين تحتاج إلى استيعاب مجال لا تعرفه. هنا، الهدف هو فهم الاتجاهات التشريعية في حقوق الملكية الفكرية وعلاقتها بالذكاء الاصطناعي — عبر دول عديدة، تقاليد قانونية متعدّدة، ووجهات نظر مختلفة لأصحاب المصلحة.

Step 1 — A first-draft prompt, written in Arabic الخطوة 1 — مسوّدة أولى للموجِّه، مكتوبة بالعربية

A subject-matter human starts with what they want — broad coverage of legislative trends, court cases, public-opinion debates, and the cultural/economic dimensions — but the draft is a single dense paragraph, not yet shaped for an AI research tool. يبدأ المختصّ بما يريده — تغطية واسعة للاتجاهات التشريعية، القضايا أمام المحاكم، نقاشات الرأي العام، والأبعاد الثقافية والاقتصادية — لكن المسوّدة فقرة كثيفة واحدة، لم تُهيَّأ بعد لأداة بحث بالذكاء الاصطناعي.

مصادر أوّلية ورئيسية عن اتجاهات تشريعية وقوانين مستحدثة وأفكار جديدة
ومقترحات في مجال حقوق الملكية الفكرية وعلاقتها بالذكاء الاصطناعي،
وتناول تعلّم الذكاء الاصطناعي، منتجات الذكاء الاصطناعي التوليدي، وكافة
القضايا المستحدثة في حقوق الملكية الفكرية والناتجة عن مستحدثات الذكاء
الاصطناعي.

نوع المصادر ليشمل فيديوهات يوتيوب عن محاضرات ومؤتمرات متخصّصة
وأكاديمية، ومواقع ومستندات عن دراسات وتشريعات واتجاهات فكرية تغطّي
دولاً مختلفة من أنحاء العالم. كما يجب عليك توفير مصادر إضافية تغطّي
قضايا أمام المحاكم أو تمّ الحكم فيها في هذا الشأن في دول مختلفة، وقضايا
رأي عام، مع عدم إهمال البُعد الاقتصادي والإنساني والثقافي والهوية.

Step 2 — Hand the draft to an LLM and ask it to optimize for NotebookLM الخطوة 2 — سلّم المسوّدة إلى نموذج لغوي واطلب منه تحسينها لتناسب NotebookLM

Instead of guessing how to structure a Deep Research brief, hand the rough English translation back to a general-purpose LLM with one explicit ask: "look up best-practice prompting for NotebookLM Deep Research, then rewrite my prompt accordingly". بدلاً من التخمين في كيفية هيكلة موجّه بحث عميق، أعِد الترجمة الأولية بالإنجليزية إلى نموذج لغوي عام مع طلب صريح: "ابحث عن أفضل ممارسات صياغة الموجّهات لـ NotebookLM Deep Research، ثم أعِد كتابة موجّهي وفقاً لها".

this is a prompt I will provide to notebooklm deep research to
help me research a topic. familiarise yourself online with best
way to prompt "notebooklm adding sources through deep research"
before optimizing my prompt:

Primary and main sources about legislative trends, new laws,
new ideas and proposals in the field of intellectual property
rights and their relationship with artificial intelligence and
addressing the learning of artificial intelligence, generative
artificial intelligence products, and all new cases in
intellectual property rights resulting from artificial
intelligence developments, the type of sources to include
YouTube videos about specialized and academic lectures and
conferences, sites and documents on studies, legislation and
intellectual trends covering different countries from around
the world. You must also provide additional sources covering
cases before the courts or have been judged in this regard in
different countries and public opinion issues without
neglecting the economic, humanitarian, cultural and identity
dimension.

Step 3 — The LLM's optimised prompt for NotebookLM Deep Research الخطوة 3 — الموجِّه المُحسَّن من النموذج اللغوي لـ NotebookLM Deep Research

The LLM returns a structured brief organised around the elements NotebookLM's Deep Research mode expects: a clear research topic, geographic scope, dimensions to cover, source types to prioritise, and explicit output goals — including instructions to surface non-Western and Global South materials. يُعيد النموذج إحاطة منظَّمة حول العناصر التي يتوقّعها وضع البحث العميق في NotebookLM: موضوع بحث واضح، نطاق جغرافي، أبعاد للتغطية، أنواع مصادر يُفضَّل التركيز عليها، وأهداف مخرجات صريحة — مع تعليمات لإبراز المواد غير الغربية ومن الجنوب العالمي.

Research topic
Legislative trends, new laws, legal proposals, and emerging
legal doctrines on intellectual property rights in relation
to artificial intelligence — with a focus on (a) AI training
and the use of copyrighted material, (b) generative AI outputs
and authorship/ownership, and (c) novel IP cases arising from
recent AI developments. Cover developments from roughly 2022
to the present.

Geographic scope
Global and comparative. Include the United States, European
Union (especially the AI Act and CDSM Directive), United
Kingdom, China, Japan, South Korea, India, Canada, Australia,
Brazil, and the Gulf/MENA region (UAE, Saudi Arabia, Egypt).
Where possible, include emerging-market and Global South
perspectives, not only Western jurisdictions.

Dimensions to cover
  - Legislation and regulation — enacted laws, draft bills,
    regulatory guidance, official consultations on AI and IP
    (copyright, patents, trademarks, trade secrets, database
    rights, sui generis rights for AI outputs, TDM exceptions).
  - Policy proposals and intellectual trends — white papers
    and reports from WIPO, OECD, UNESCO, U.S. Copyright Office,
    European Commission, UK IPO, CNIPA, JPO, etc.; academic
    proposals for new legal frameworks.
  - Court cases and litigation — pending and decided cases on
    AI training, output infringement, and authorship (e.g.
    NYT v. OpenAI, Getty v. Stability AI, Andersen v. Stability
    AI, Thaler / DABUS, Bartz v. Anthropic, Chinese AIGC
    rulings, Japanese guidance cases).
  - Public opinion and civil society — surveys, artist and
    creator coalitions, open letters, position statements.
  - Economic dimension — impact on creative industries,
    licensing markets, the music/publishing/visual-arts
    economy, labour displacement, competition concerns.
  - Humanitarian, cultural, and identity dimensions — effects
    on indigenous knowledge, traditional cultural expressions,
    minority languages, cultural heritage, moral rights, and
    voice/likeness/right-of-publicity issues (incl. deepfakes).

Source types to prioritise (a balanced mix)
  - Primary legal sources: statutes, draft bills, court
    opinions, regulatory guidance, official consultations.
  - Institutional reports: WIPO, OECD, UNESCO, national
    copyright/patent offices, parliamentary research services.
  - Academic sources: peer-reviewed law reviews, SSRN papers,
    university research centres (Stanford CIS, Oxford OII,
    Max Planck).
  - YouTube videos: recorded lectures, university symposia,
    bar association panels, WIPO/Berkman Klein/CREATe talks,
    parliamentary or congressional hearings.
  - Quality journalism: legal industry publications (IPKat,
    JIPLP, Kluwer Copyright Blog, Lexology, Law360) and
    major-outlet investigative reporting.
  - Civil society and industry: position papers from creator
    unions, publisher and record-label associations, AI labs,
    EFF/Creative Commons-style organisations.

Output goals
  - A comprehensive, diverse source library spanning all six
    dimensions and the geographies above.
  - Avoid over-concentration on US/EU sources; deliberately
    surface non-English and Global South materials.
  - Prefer primary and authoritative sources over secondary
    commentary when both exist.
  - Include sources representing competing viewpoints
    (rights-holders, AI developers, users, governments,
    academics).

Step 4 — What NotebookLM produced from those sources الخطوة 4 — ما أنتجه NotebookLM من تلك المصادر

Audio Audio overview — Arabic ملخّص صوتي — بالعربية
IP & AI mind map — NotebookLM output
Image Mind map of the field خريطة ذهنية للمجال
IP & AI infographic brief — NotebookLM output
Image One-page brief إحاطة من صفحة واحدة

Step 5 — Take the notebook into Gemini and write الخطوة 5 — انقل الدفتر إلى Gemini وابدأ الكتابة

From here, the same NotebookLM notebook becomes a grounded knowledge base inside Gemini — every claim in the article you draft can be traced back to a source you've already curated. The model writes; you supervise. من هذه النقطة، يصبح دفتر NotebookLM نفسه قاعدة معرفة موثّقة داخل Gemini — كل ادّعاء في المقال الذي تكتبه يمكن تتبّعه إلى مصدر اخترته بنفسك مسبقاً. النموذج يكتب، وأنت تُشرف.

The shift: you stopped guessing the right prompt and started composing with the AI — one model to write better prompts, another to research, a third to draft. Each used for what it does best. التحوّل: توقّفتَ عن التخمين في الموجِّه المناسب وبدأت في التأليف مع الذكاء الاصطناعي — نموذج لكتابة موجّهات أفضل، آخر للبحث، وثالث للصياغة. كل واحد مُستخدَم في ما يجيده.
Block I · Topic B · Brand identity المحور I · الموضوع ب · هوية العلامة

Wheeler's brand identity framework. إطار Wheeler لهوية العلامة التجارية.

Unit وحدة I.1

What is brand identity? ما هي هوية العلامة التجارية؟

Brand identity is the tangible expression of a brand — everything you can see, touch, hear, and experience. It answers: how do we look, how do we sound, and what do we stand for consistently? هوية العلامة التجارية هي التعبير الملموس عن العلامة — كل ما يمكنك رؤيته، لمسه، سماعه، وتجربته. تجيب عن: كيف نبدو، كيف نتحدث، وما الذي نمثّله باستمرار؟

"Brand identity is the visual and verbal expression of a brand. It starts with a name and a symbol." — Alina Wheeler "هوية العلامة التجارية هي التعبير البصري واللفظي عنها. تبدأ باسم ورمز." — ألينا ويلر
Unit وحدة I.2

Wheeler's core elements عناصر Wheeler الأساسية

Elementالعنصر What it meansما يعنيه What to look forما الذي نبحث عنه
Brand voiceصوت العلامة How the brand speaks and writesكيف تتحدث العلامة وتكتب Formal/casual, emotional/rationalرسمي/غير رسمي، عاطفي/منطقي
Messagingالرسائل الأساسية Core ideas and promises repeatedالأفكار والوعود المتكررة Taglines, recurring themesالشعارات، المحاور المتكررة
Visual identityالهوية البصرية Consistent look across all touchpointsالمظهر المتسق عبر كل نقاط التواصل Colors, fonts, image styleالألوان، الخطوط، أسلوب الصور
Name & taglineالاسم والشعار The verbal anchor of the brandالمرساة اللفظية للعلامة Clarity, memorabilityالوضوح، سهولة التذكر
Brand storyقصة العلامة The narrative behind the brandالرواية خلف العلامة Origin, mission, visionالنشأة، الرسالة، الرؤية
Valuesالقيم Non-negotiable beliefsالمبادئ غير القابلة للتفاوض What they stand for publiclyما تدعمه العلامة علنًا
Target audienceالجمهور المستهدف Who they speak toمن تخاطبه العلامة Language, references, examples usedاللغة، المراجع، الأمثلة المستخدمة
Unit وحدة I.3

What we extract from competitor posts ما الذي نستخلصه من منشورات المنافسين؟

Wheeler elementعنصر Wheeler Signal in social media postsالإشارة في منشورات وسائل التواصل
Brand voiceصوت العلامةWord choice, sentence length, punctuation styleاختيار الكلمات، طول الجمل، أسلوب علامات الترقيم
Messagingالرسائل الأساسيةRepeated phrases, content pillarsالعبارات المتكررة، محاور المحتوى
Visual identityالهوية البصريةColor palette, design templates, photo styleلوحة الألوان، قوالب التصميم، أسلوب الصور
Brand storyقصة العلامةBio references, origin mentionsالإشارات في السيرة الذاتية، ذكر النشأة
ValuesالقيمCauses supported, positions takenالقضايا المدعومة، المواقف المُعلنة
Target audienceالجمهور المستهدفWho they tag, who engages, language registerمن يُذكرون، من يتفاعل، مستوى اللغة
Unit وحدة I.4

Why this matters for you لماذا يهمك هذا؟

  • Highly practical — built for real-world application.عملي للغاية — مبني للتطبيق الفعلي في الواقع.
  • Maps directly to what you actually produce — copy, visuals, campaigns.يرتبط مباشرة بما تنتجه فعلاً — النصوص، البصريات، الحملات.
  • Makes it easy to brief both humans and AI on your brand.يسهّل توجيه كل من البشر والذكاء الاصطناعي حول علامتك.
Block I · Topic C · Message house المحور I · الموضوع ج · بيت الرسالة

One blueprint for your brand's voice. مخطط واحد لصوت علامتك.

Unit وحدة I.5

What is a message house? ما هو بيت الرسالة؟

A message house is a strategic communication framework that organizes everything a brand says into one coherent, consistent structure. It ensures that no matter who speaks, what channel is used, or what audience is addressed — the brand always sounds like itself. بيت الرسالة هو إطار استراتيجي للتواصل يُنظّم كل ما تقوله العلامة في بنية واحدة متماسكة ومتسقة. يضمن أنه بغض النظر عن من يتحدث، أو القناة المستخدمة، أو الجمهور المُخاطَب — تبدو العلامة دائمًا كما هي.

Think of it as the blueprint of your brand's voice. فكّر فيه كـ مخطط صوت علامتك التجارية.
Unit وحدة I.6

The structure الهيكل

Roofالسقف
The core message — one line everyone remembers. الرسالة الجوهرية — جملة واحدة يتذكّرها الجميع.
Pillar 1الركيزة 1
Supporting messageرسالة داعمة
Pillar 2الركيزة 2
Supporting messageرسالة داعمة
Pillar 3الركيزة 3
Supporting messageرسالة داعمة
Foundationالأساس
Brand values & positioning statement. قيم العلامة وبيان التموضع.
Unit وحدة I.7

Each part explained شرح كل جزء

Partالجزء Roleالدور Exampleمثال
Roofالسقف The single most important thing you want people to rememberالشيء الواحد الأهم الذي تريد الناس أن يتذكروه "We make complex technology feel human""نجعل التكنولوجيا المعقدة تبدو إنسانية"
Pillarsالركائز 3 key messages that support and prove the roof3 رسائل رئيسية تدعم وتُثبت السقف Innovation / Simplicity / Trustالابتكار / البساطة / الثقة
Foundationالأساس The values, mission, and positioning that everything rests onالقيم والرسالة والتموضع الذي يرتكز عليه كل شيء "We believe technology should serve people, not the other way around""نؤمن بأن التكنولوجيا يجب أن تخدم الإنسان"
Unit وحدة I.8

How to build your message house كيف تبني بيت رسالتك؟

01
Lay the foundation ضع الأساس

Define your brand mission, your core values, and your positioning statement — who you are, for whom, and why you're different. حدّد رسالة علامتك، قيمك الجوهرية، وبيان تموضعك — من أنت، لمن، ولماذا أنت مختلف.

02
Raise the pillars أقم الركائز

Choose three key messages that support the core, can each be proven with evidence or stories, and cover what your audience most needs to believe. اختر ثلاث رسائل رئيسية داعمة لرسالتك الجوهرية، قابلة للإثبات بأدلة أو قصص، ومُغطّية لأهم ما يحتاج جمهورك إلى تصديقه.

03
Build the roof ابنِ السقف

Write one sentence that captures the entire promise — memorable, repeatable, and usable in any context: pitch, post, campaign, or conversation. اكتب جملة واحدة تُلخّص الوعد بالكامل — سهلة التذكر والتكرار وصالحة في أي سياق: عرض، منشور، حملة، أو محادثة.

Unit وحدة I.9

Message house in action — example بيت الرسالة في التطبيق — مثال

Layerالطبقة Contentالمحتوى
Roofالسقف"Empowering Arab professionals to lead with confidence""تمكين المحترفين العرب من القيادة بثقة"
Pillar 1الركيزة 1We provide cutting-edge skills trainingنقدّم تدريبًا على أحدث المهارات
Pillar 2الركيزة 2We speak your language and understand your marketنتحدث لغتك ونفهم سوقك
Pillar 3الركيزة 3We build community, not just coursesنبني مجتمعًا، لا مجرد دورات
FoundationالأساسRooted in a belief that Arab talent is world-class and deserves world-class toolsإيمان راسخ بأن المواهب العربية عالمية المستوى وتستحق أدوات عالمية
Unit وحدة I.10

The AI connection صلة الذكاء الاصطناعي

When you give AI your message house, it can: عندما تزود الذكاء الاصطناعي ببيت رسالتك، يستطيع:

  • Generate on-brand content across all formats.إنتاج محتوى يعكس هويتك عبر جميع الصيغ.
  • Maintain consistent tone across 100 posts.الحفاظ على نبرة متسقة عبر 100 منشور.
  • Flag content that contradicts your messaging.تحديد المحتوى الذي يتعارض مع رسائلك.
  • Build entire content calendars aligned to your pillars.بناء تقويمات محتوى كاملة مرتبطة بركائزك.
Workshop block II · 20 minutes المحور الثاني من الورشة · 20 دقيقة

CXJM — customer experience journey mapping. CXJM — رسم خرائط رحلة تجربة العميل.

Two topics, fourteen units total. We define the people we are trying to reach — personas — then look at how surveys feed those personas, and how AI changes what surveys can do. موضوعان، أربع عشرة وحدة إجمالاً. نُعرّف الأشخاص الذين نحاول الوصول إليهم — الشخصيات — ثم نبحث في كيفية تغذية الاستبيانات لها، وكيف يُغيّر الذكاء الاصطناعي ما يمكن للاستبيانات فعله.

Block II · Topic A · Personas المحور II · الموضوع أ · الشخصيات

Customer personas — built for connection. شخصيات العملاء — مبنيّة للتواصل.

Unit وحدة II.1

What is a customer persona? ما هي شخصية العميل؟

A customer persona is a semi-fictional representation of your ideal customer, built from real data, research, and informed assumptions. It is not a demographic profile — it is a living portrait of the human being you are trying to reach. شخصية العميل هي تمثيل شبه خيالي لعميلك المثالي، مبني من بيانات حقيقية وأبحاث وافتراضات مدروسة. هي ليست مجرد ملف ديموغرافي — هي صورة حية للإنسان الذي تحاول الوصول إليه.

"You cannot connect with 'everyone'. You can connect deeply with someone." "لا يمكنك التواصل مع الجميع. يمكنك التواصل بعمق مع شخص واحد."
Unit وحدة II.2

Why personas matter لماذا تهم الشخصيات؟

Without a personaبدون شخصية With a personaمع شخصية
You write for everyoneتكتب للجميعYou write for one personتكتب لشخص واحد
Content feels genericالمحتوى يبدو عامًاContent feels personalالمحتوى يبدو شخصيًا
Hard to choose topicsصعوبة اختيار المواضيعTopics become obviousالمواضيع تصبح واضحة
AI generates average outputالذكاء الاصطناعي ينتج مخرجات متوسطةAI generates targeted outputالذكاء الاصطناعي ينتج مخرجات موجّهة
Unit وحدة II.3

Core elements of a persona العناصر الجوهرية للشخصية

Elementالعنصر What to defineما يجب تحديده
Name & photoالاسم والصورةGive them a human identityأعطها هوية إنسانية
DemographicsالديموغرافياAge, gender, location, education, incomeالعمر، الجنس، الموقع، التعليم، الدخل
Job & roleالوظيفة والدورWhat they do, their responsibilitiesما يفعله، مسؤولياته
GoalsالأهدافWhat they are trying to achieveما يسعى لتحقيقه
Pain pointsنقاط الألمWhat frustrates or blocks themما يُحبطه أو يعيقه
ValuesالقيمWhat they believe inما يؤمن به
Media habitsعادات الإعلامWhere they consume contentأين يستهلك المحتوى
Buying behaviorسلوك الشراءHow they make decisionsكيف يتخذ القرارات
ChannelsالقنواتWhere and how to reach them — online, offline, and socialأين وكيف تصل إليه — أونلاين، أوفلاين، وسائل التواصل
QuoteالاقتباسOne sentence they would say about their lifeجملة واحدة يقولها عن حياته
Unit وحدة II.4

Channel types to consider أنواع القنوات التي يجب مراعاتها

Channel typeنوع القناة Examplesأمثلة
Social mediaوسائل التواصل الاجتماعيLinkedIn, Instagram, TikTok, X, Facebook, Snapchatلينكدإن، إنستغرام، تيك توك، إكس، فيسبوك، سناب شات
OnlineأونلاينEmail newsletters, Google Search, YouTube, podcasts, blogsالنشرات البريدية، بحث Google، يوتيوب، البودكاست، المدونات
PhysicalفيزيائيةEvents, conferences, retail stores, word of mouthالفعاليات، المؤتمرات، المتاجر، التوصيات الشفهية
MobileموبايلWhatsApp, SMS, push notifications, appsواتساب، رسائل SMS، الإشعارات، التطبيقات
CommunityمجتمعيةForums, WhatsApp groups, industry associationsالمنتديات، مجموعات واتساب، الجمعيات المهنية
Unit وحدة II.5

A persona in practice — "Ambitious Ahmed" شخصية في التطبيق — "أحمد الطموح"

Ageالعمر32
RoleالدورMarketing Manager at a mid-size UAE companyمدير تسويق في شركة إماراتية متوسطة الحجم
GoalالهدفProve ROI on digital campaigns to leadershipإثبات العائد على الاستثمار للحملات الرقمية أمام القيادة
Pain pointنقطة الألمToo much data, not enough insightبيانات كثيرة جدًا، ورؤى قليلة جدًا
ValuesالقيمGrowth, efficiency, being seen as innovativeالنمو، الكفاءة، أن يُنظر إليه كمبتكر
MediaالإعلامLinkedIn, podcasts, YouTube tutorialsلينكدإن، البودكاست، دروس يوتيوب
Buying behaviorسلوك الشراءResearches extensively before deciding, trusts peer reviewsيبحث بعمق قبل القرار، يثق بآراء الأقران
Quoteالاقتباس"I know AI can help me, I just don't know where to start.""أعرف أن الذكاء الاصطناعي يمكنه مساعدتي، لكنني لا أعرف من أين أبدأ."
Unit وحدة II.6

Ahmed's channel map خريطة قنوات أحمد

Channel typeنوع القناة Where to find Ahmedأين تجد أحمد How to reach himكيف تصل إليه
Social mediaوسائل التواصلLinkedIn (daily), Instagram (casual browsing)لينكدإن (يوميًا)، إنستغرام (تصفح عشوائي)Thought leadership posts, case studies, carouselsمنشورات قيادة فكرية، دراسات حالة، كاروسيل
OnlineأونلاينGoogle Search, industry newsletters, YouTubeبحث Google، نشرات القطاع، يوتيوبSEO content, email sequences, tutorial videosمحتوى SEO، تسلسلات بريد إلكتروني، فيديوهات تعليمية
MobileموبايلWhatsApp professional groupsمجموعات واتساب المهنيةShared articles, community presenceمشاركة المقالات، الحضور المجتمعي
PhysicalفيزيائيةIndustry conferences, business breakfastsمؤتمرات القطاع، إفطارات الأعمالSpeaking engagements, networking, branded materialsجلسات تحدّث، تواصل، مواد مرئية
CommunityمجتمعيةMarketing & tech forums, Slack groupsمنتديات التسويق والتقنية، مجموعات SlackExpert contributions, Q&A participationمساهمات الخبراء، المشاركة في الأسئلة
Unit وحدة II.7

How AI builds channel maps for you كيف يبني الذكاء الاصطناعي خرائط القنوات لك؟

Feed AIما تُقدّمه
  • Persona descriptionوصف الشخصية
  • Industry contextسياق الصناعة
  • Geographic marketالسوق الجغرافي
Ask AI to returnما تطلبه منه
  • Most effective channels ranked by reachأكثر القنوات فعالية مرتبة حسب الوصول
  • Best content format per channelأفضل صيغة محتوى لكل قناة
  • Optimal posting frequency per channelالتكرار الأمثل للنشر لكل قناة
  • Tone adjustment per channelتعديل النبرة حسب كل قناة
Unit وحدة II.8

Persona → channel → content flow التدفق الكامل: شخصية ← قناة ← محتوى

Persona pain pointنقطة ألم الشخصية
Identify relevant channelsتحديد القنوات المناسبة
Match format to channelمطابقة الصيغة مع القناة
AI prompt with persona + channel contextموجّه ذكاء اصطناعي مع سياق الشخصية والقناة
Right message · Right person · Right placeالرسالة الصحيحة · للشخص الصحيح · في المكان الصحيح
Block II · Topic B · Surveys المحور II · الموضوع ب · الاستبيانات

Surveys, reimagined with AI. الاستبيانات بأعين الذكاء الاصطناعي.

Unit وحدة II.9

What surveys can and cannot do ما تستطيع الاستبيانات فعله وما لا تستطيعه

A survey is a blunt instrument with precision potential — only if you design it right. Most marketers expect too much from a single instrument, or design surveys that answer the wrong questions entirely. الاستبيان هو أداة غير دقيقة بإمكانية دقة عالية — لكن فقط إذا صممته بشكل صحيح. معظم المسوّقين يتوقعون الكثير من أداة واحدة، أو يصممون استبيانات تجيب على أسئلة خاطئة تمامًا.

The attention economy problem مشكلة اقتصاد الانتباه

Survey lengthطول الاستبيان Completion rateمعدل الإكمال Recommended useالاستخدام الموصى به
1–3 questions1–3 أسئلة80–90%Pulse checks, NPS, single-topic feedbackفحوصات نبض، NPS، تغذية راجعة على موضوع واحد
5–7 questions5–7 أسئلة60–70%Focused persona or behavior snapshotsلقطات مركّزة للشخصية أو السلوك
10–12 questions10–12 سؤالاً40–55%Comprehensive persona mapping — maximum for general audiencesرسم خرائط شاملة للشخصية — الحد الأقصى للجماهير العامة
15–20 questions15–20 سؤالاً20–35%Only for highly motivated respondentsللمستجيبين ذوي الدوافع العالية فقط
20+ questionsأكثر من 20 سؤالاًBelow 20%أقل من 20%Academic research onlyبحث أكاديمي فقط
The rule: design for the respondent's patience, not your curiosity. القاعدة: صمم بناءً على صبر المستجيب، لا فضولك.

The best question mix for persona + behavior research أفضل مزيج من الأسئلة لبحث الشخصية والسلوك

For a 10-question persona/behavior survey, this is the proven mix: لاستبيان شخصية/سلوك من 10 أسئلة، هذا هو المزيج المُثبت:

Question typeنوع السؤال How manyالعدد What it gives youما يعطيك
Demographicديموغرافي2–3Persona segmentation anchorsمراسي تجزئة الشخصية
Behavioralسلوكي3Actual patterns, not assumptionsأنماط فعلية لا افتراضات
Attitudinalموقفي2Motivation and values layerطبقة الدوافع والقيم
Open-endedمفتوح1–2Language, emotion, unexpected insightاللغة، العاطفة، الرؤى غير المتوقعة
Ranking / priorityترتيب / أولوية1Forces real tradeoffsيفرض مقايضات حقيقية
Never lead with open-ended questions — put them at question 8 or 9. لا تبدأ أبدًا بأسئلة مفتوحة — ضعها في السؤال 8 أو 9.

Tradeoffs every marketer must understand المقايضات التي يجب على كل مسوّق فهمها

Tradeoffالمقايضة The realityالواقع
Scale vs. depthالحجم مقابل العمقMore respondents = more confidence in patterns. Fewer respondents with richer questions = better persona texture. You rarely get both.المزيد من المستجيبين = ثقة أكبر في الأنماط. مستجيبون أقل بأسئلة أغنى = نسيج شخصية أفضل. نادرًا ما تحصل على كليهما.
Closed vs. openالمغلقة مقابل المفتوحةClosed questions give you numbers fast. Open questions give you truth slowly. AI is changing the "slowly" part dramatically.الأسئلة المغلقة تعطيك أرقامًا بسرعة. الأسئلة المفتوحة تعطيك الحقيقة ببطء. الذكاء الاصطناعي يغيّر جزء "البطء" الآن.
Broad vs. focusedالواسع مقابل المركزA survey that tries to answer 5 business questions answers none well. One survey, one objective.استبيان يحاول الإجابة على 5 أسئلة تجارية لا يجيب على أي منها جيدًا. هدف واحد فقط.
Anonymous vs. namedمجهول مقابل مسمىAnonymous = more honest answers. Named = ability to follow up. For behavior research, anonymous wins almost always.مجهول = إجابات أكثر صدقًا. مسمى = القدرة على المتابعة. للسلوك، المجهول يفوز دائمًا تقريبًا.
Frequency vs. fatigueالتكرار مقابل الإرهاقSurveying the same audience more than once per quarter risks response fatigue and lower-quality answers.استطلاع الجمهور أكثر من مرة في الربع يخاطر بإرهاق الاستجابة وجودة أقل.

Do's and don'ts افعل ولا تفعل

Doافعل
  • Define ONE primary objective before writing a single question.حدّد هدفًا أساسيًا واحدًا قبل كتابة أي سؤال.
  • Test your survey on 3–5 people before launch.اختبر استبيانك على 3–5 أشخاص قبل الإطلاق.
  • Use behavioral questions before attitudinal ones.استخدم الأسئلة السلوكية قبل الموقفية.
  • Include one forced-ranking question.ضمّن سؤال ترتيب إجباري واحدًا.
  • Tell respondents exactly how long it takes.أخبر المستجيبين بالمدة الدقيقة.
  • Use skip logic for irrelevant segments.استخدم منطق التخطي للشرائح غير ذات الصلة.
Don'tلا تفعل
  • Start writing questions before defining the decision the survey must support.ابدأ في كتابة الأسئلة قبل تحديد القرار الذي يدعمه الاستبيان.
  • Launch to your full list and discover broken logic mid-campaign.أطلقه لقائمتك الكاملة واكتشف منطقًا معطوبًا لاحقًا.
  • Lead with opinions — behavior data grounds everything else.ابدأ بالآراء — بيانات السلوك تؤسّس كل شيء آخر.
  • Give them 10 checkboxes and call it "priority research".أعطهم 10 خانات اختيار وتُسمّيها "بحث الأولويات".
  • Let them discover mid-survey it's longer than expected.دعهم يكتشفون في منتصف الاستبيان أنه أطول مما توقعوا.
  • Make every respondent answer every question.اجعل كل مستجيب يجيب على كل سؤال.
Unit وحدة II.10

Designing for persona & behavior extraction التصميم لاستخلاص الشخصية والسلوك

The most common mistake: designing a survey to confirm what you already believe, rather than to discover what you don't know. الخطأ الأكثر شيوعًا: تصميم استبيان لتأكيد ما تؤمن به بالفعل، بدلاً من اكتشاف ما لا تعرفه.

Question architecture for persona extraction هندسة الأسئلة لاستخلاص الشخصية

Persona elementعنصر الشخصية Question typeنوع السؤال Example questionمثال على السؤال
DemographicsالديموغرافياClosed, single choiceمغلق، اختيار واحدWhat is your current role?ما هو دورك الحالي؟
GoalsالأهدافRankingترتيبRank your top 3 professional priorities this year.رتّب أهم 3 أولويات مهنية لهذا العام.
Pain pointsنقاط الألمOpen-ended or multi-selectمفتوح أو متعدد الاختيارWhat is the single biggest obstacle to achieving your goals?ما هو العائق الأكبر أمام تحقيق أهدافك؟
ValuesالقيمAttitudinal scaleمقياس موقفيHow important is brand reputation vs. price when choosing a vendor?ما مدى أهمية سمعة العلامة مقابل السعر عند اختيار مورد؟
Media habitsعادات الإعلامBehavioral, multi-selectسلوكي، متعدد الاختيارWhich platforms do you use weekly for professional content?ما المنصات التي تستخدمها أسبوعيًا للمحتوى المهني؟
ChannelsالقنواتBehavioral, frequencyسلوكي، تكرارHow do you prefer to receive industry updates?كيف تفضل تلقي تحديثات القطاع؟
Buying behaviorسلوك الشراءScenario-basedقائم على سيناريوWhen evaluating a new tool, what do you do first?عند تقييم أداة جديدة، ماذا تفعل أولاً؟
Language & voiceاللغة والصوتOpen-endedمفتوحDescribe your biggest challenge in your own words.صف أكبر تحدياتك بكلماتك الخاصة.

The segmentation question — your most valuable سؤال التجزئة — أثمن أسئلتك

Always include one question whose sole purpose is to split your respondents into groups. دائمًا ضمّن سؤالاً واحدًا هدفه الوحيد تقسيم المستجيبين إلى مجموعات.

  • "How long have you been in your current role?" → splits by seniority."كم من الوقت أمضيت في دورك الحالي؟" ← يُقسّم حسب الأقدمية.
  • "What is the primary size of your organization?" → splits by context."ما هو الحجم الأساسي لمنظمتك؟" ← يُقسّم حسب السياق.
  • "How would you describe your current use of AI tools?" → splits by adoption stage."كيف تصف استخدامك الحالي لأدوات الذكاء الاصطناعي؟" ← يُقسّم حسب مرحلة التبنّي.
Once you have segments, every other answer becomes 4× more useful. بمجرد حصولك على شرائح، كل إجابة أخرى تصبح أكثر فائدة بـ 4 مرات.

Behavioral questions: the underused gold mine الأسئلة السلوكية: المنجم الذهبي غير المستغل

Weak attitudinalموقفي ضعيف Strong behavioral replacementبديل سلوكي قوي
"How important is social media to your work?""ما مدى أهمية وسائل التواصل لعملك؟""How many hours per week do you spend on social media for professional purposes?""كم ساعة في الأسبوع تقضي على وسائل التواصل لأغراض مهنية؟"
"Do you value continuous learning?""هل تقدّر التعلم المستمر؟""When did you last complete a professional course or attend an industry event?""متى أكملت آخر دورة مهنية أو حضرت آخر فعالية قطاع؟"
"Is budget a concern for you?""هل الميزانية مصدر قلق لك؟""In the last 12 months, how many tools or services did you cancel due to cost?""في الأشهر الـ 12 الماضية، كم أداة أو خدمة ألغيت اشتراكها بسبب التكلفة؟"
Unit وحدة II.11

From results to persona profiles من النتائج إلى ملفات الشخصية

Getting 300 responses back is not a persona — it is a dataset. The gap between the two is where most marketers stop, and where AI now gives you an enormous advantage. الحصول على 300 إجابة ليس شخصية — إنه مجموعة بيانات. الفجوة بين الاثنين هي حيث يتوقف معظم المسوّقين، وحيث يمنحك الذكاء الاصطناعي الآن ميزة هائلة.

The classic manual flow (and why it breaks) التدفق اليدوي الكلاسيكي (ولماذا يفشل)

300 responses collected300 إجابة محصّلة
Export to spreadsheetتصدير إلى جدول بيانات
Filter by demographic segmentsتصفية حسب الشرائح الديموغرافية
Calculate averages on closed questionsحساب المتوسطات على الأسئلة المغلقة
Read open-ended answers manually — this is where it diesقراءة الإجابات المفتوحة يدويًا — هنا يموت كل شيء
Write a summary report nobody readsكتابة تقرير ملخّص لا يقرأه أحد
Persona based on averages — this persona is fictionalشخصية مبنية على المتوسطات — هذه الشخصية خيالية
No real human is average. Averages flatten the exact differences that make personas useful. لا يوجد إنسان حقيقي متوسط. المتوسطات تُسطّح الاختلافات التي تجعل الشخصيات مفيدة.

What you actually need: cluster-based personas ما تحتاجه فعلاً: شخصيات قائمة على التجميع

Cluster signalإشارة التجميع What to look forما تبحث عنه
Behavioral clusterتجميع سلوكيPeople who do the same things in the same sequenceأشخاص يفعلون نفس الأشياء بنفس التسلسل
Goal clusterتجميع أهدافPeople who share the same primary objectiveأشخاص يشتركون في نفس الهدف الأساسي
Barrier clusterتجميع حواجزPeople blocked by the same obstacleأشخاص يعيقهم نفس العائق
Channel clusterتجميع قنواتPeople who consume content exclusively in the same placesأشخاص يستهلكون المحتوى في نفس الأماكن حصرًا
When 3 or more of these overlap in the same group — you have found a real persona. عندما تتداخل 3 أو أكثر من هذه في نفس المجموعة — لقد وجدت شخصية حقيقية.

Reading the gaps: what surveys can't tell you قراءة الفجوات: ما لا يستطيع الاستبيان إخبارك به

Surveys tell you wellالاستبيانات تخبرك جيدًا Surveys cannot tell youالاستبيانات لا تخبرك
What people do and how oftenما يفعله الناس وكم مرةWhy they really do itلماذا يفعلونه حقًا
What they say they preferما يقولون إنهم يفضلونهWhat they actually choose under pressureما يختارونه فعلاً تحت الضغط
Surface-level pain pointsنقاط الألم السطحيةThe emotional weight behind the painالثقل العاطفي وراء الألم
General channel preferencesتفضيلات القناة العامةThe specific moment they are most receptiveاللحظة المحددة التي يكونون فيها الأكثر تقبلاً
Unit وحدة II.12

The intelligence stack مجموعة الاستخبارات

One survey is a snapshot. A stack is a story. Each new layer sharpens and challenges what came before. استبيان واحد هو لقطة. المجموعة هي قصة. كل طبقة جديدة تُصقل وتُتحدى ما قبلها.

  1. L1
    Survey (scale)الاستبيان (الحجم)

    Broad patterns, segmentation, behavioral baselines. 100–500+ respondents · low depth · high confidence in patterns.أنماط واسعة، تجزئة، خطوط أساسية سلوكية. 100–500+ مستجيب · عمق منخفض · ثقة عالية في الأنماط.

  2. L2
    Targeted follow-up survey (focus)استبيان مستهدف للمتابعة (التركيز)

    Sent only to specific segments from L1. 20–50 respondents · medium depth · validates or challenges L1.يُرسل فقط لشرائح محددة من L1. 20–50 مستجيب · عمق متوسط · يتحقق أو يتحدى L1.

  3. L3
    Focus groups (depth)مجموعات التركيز (العمق)

    3–6 people from your most important persona segment. Explore the WHY behind L1 and L2.3–6 أشخاص من شريحة شخصيتك الأهم. استكشف "لماذا" وراء L1 و L2.

  4. L4
    Customer journey touchpoint dataبيانات نقاط تماس رحلة العميل

    Real behavior — analytics, email opens, purchases, support tickets, social engagement.السلوك الحقيقي — التحليلات، فتح البريد، المشتريات، تذاكر الدعم، التفاعل الاجتماعي.

  5. L5
    Continuous listeningالاستماع المستمر

    Social listening, comment analysis, DM patterns, community conversations — ongoing, not periodic.الاستماع الاجتماعي، تحليل التعليقات، أنماط الرسائل المباشرة، محادثات المجتمع — مستمر، لا دوري.

When to use each layer متى تستخدم كل طبقة

Layerالطبقة Use whenاستخدمها عندما Outputالمخرج
Broad surveyالاستبيان الواسعNeed to understand audience at scaleتحتاج فهم جمهورك على نطاق واسعPersona hypothesesفرضيات الشخصية
Targeted surveyالاستبيان المستهدفNeed to validate a specific assumptionتحتاج التحقق من افتراض محددConfirmed or rejected assumptionافتراض مؤكد أو مرفوض
Focus groupمجموعة التركيزNeed emotional depth and languageتحتاج عمقًا عاطفيًا ولغةPersona voice, real quotesصوت الشخصية، اقتباسات حقيقية
Journey dataبيانات الرحلةNeed to know what they actually do vs. sayتحتاج معرفة ما يفعلونه فعلاً مقابل ما يقولونهBehavioral reality checkفحص الواقع السلوكي
Continuous listeningالاستماع المستمرNeed to stay current as audience evolvesتحتاج البقاء حديثًا مع تطور الجمهورLiving persona updatesتحديثات شخصية حية
Unit وحدة II.13

AI transforms survey analysis الذكاء الاصطناعي يحوّل تحليل الاستبيانات

For years, open-ended survey questions were a trap. You included them because you wanted real human insight; then 400 people answered and you had no scalable way to process the responses. لسنوات، كانت أسئلة الاستبيان المفتوحة فخًا. أدرجتها لأنك أردت رؤى إنسانية حقيقية؛ ثم أجاب عليها 400 شخص ولم تكن لديك طريقة قابلة للتطوير لمعالجة الإجابات.

AI eliminates this problem entirely. الذكاء الاصطناعي يلغي هذه المشكلة تمامًا.

What AI does with open-ended responses ما يفعله الذكاء الاصطناعي بالإجابات المفتوحة

01
Classificationالتصنيف

AI reads every response and assigns it to a category — without you defining the categories first. It finds the categories in the data.يقرأ الذكاء الاصطناعي كل إجابة ويضعها في فئة — دون أن تحدد الفئات مسبقًا. يجد الفئات في البيانات.

02
Clusteringالتجميع

AI groups similar responses together — even when respondents used completely different words to express the same idea.يجمع الذكاء الاصطناعي الإجابات المتشابهة — حتى عندما يستخدم المستجيبون كلمات مختلفة تمامًا للتعبير عن نفس الفكرة.

03
Quantificationالتحويل إلى أرقام

What was qualitative becomes countable.ما كان نوعيًا يصبح قابلاً للعد.

Theme identified by AIالموضوع الذي حدّده الذكاء الاصطناعي Response countعدد الإجابات % of total% من الإجمالي
"Not enough time to learn new tools""لا وقت كافٍ لتعلم أدوات جديدة"8729%
"Leadership doesn't support AI adoption""القيادة لا تدعم تبني الذكاء الاصطناعي"6321%
"Don't know where to start""لا أعرف من أين أبدأ"5819%
"Privacy and data concerns""مخاوف الخصوصية والبيانات"4114%
"Budget limitations""قيود الميزانية"3411%
Otherأخرى176%

AI flagging: immediate action items تنبيهات الذكاء الاصطناعي: عناصر عمل فورية

Flag typeنوع التنبيه Example responseمثال على الإجابة Recommended actionالإجراء الموصى به
Urgent complaintشكوى عاجلة"I've been trying to cancel for 3 months and no one responds.""أحاول الإلغاء منذ 3 أشهر ولا أحد يرد."Immediate customer service escalationتصعيد فوري لخدمة العملاء
Churn signalإشارة مغادرة"I'm evaluating your competitors right now.""أقيّم منافسيك الآن."Sales or retention team alertتنبيه فريق المبيعات أو الاحتفاظ
Product insightرؤية منتج"I would pay more if you added X feature.""سأدفع أكثر إذا أضفت ميزة X."Product team briefingإحاطة فريق المنتج
Advocacy opportunityفرصة مناصرة"This tool changed how my entire team works.""هذه الأداة غيّرت طريقة عمل فريقي بأكمله."Case study or testimonial outreachالتواصل لدراسة حالة أو شهادة
Crisis signalإشارة أزمةMultiple respondents mention the same specific failureمستجيبون متعددون يذكرون نفس الفشل المحددImmediate internal reviewمراجعة داخلية فورية

A practical AI workflow for survey analysis سير عمل عملي للذكاء الاصطناعي لتحليل الاستبيان

Export raw survey responses (CSV or paste).
        ↓
Feed into Claude or Gemini with this prompt:
  "Here are [N] open-ended survey responses.
   1. Identify the main themes without me defining them.
   2. Group responses under each theme.
   3. Count responses per theme and give me a percentage.
   4. Flag any responses requiring immediate action.
   5. Identify which persona segment each theme belongs to."
        ↓
Review AI output — validate, adjust, challenge.
        ↓
Feed confirmed themes back into AI for
persona mapping and content strategy.
Unit وحدة II.14

AI-powered persona detection اكتشاف الشخصية بالذكاء الاصطناعي

The traditional approach built personas after weeks of manual analysis. AI compresses this to hours — and produces more nuanced results because it processes the full dataset rather than a human-readable sample. النهج التقليدي بنى الشخصيات بعد أسابيع من التحليل اليدوي. الذكاء الاصطناعي يضغط هذا إلى ساعات — ويُنتج نتائج أكثر دقة لأنه يعالج مجموعة البيانات الكاملة.

What you feed AI ما تُقدّمه للذكاء الاصطناعي

Inputالمدخل Why it mattersلماذا يهم
Full response export per respondentتصدير الإجابات الكامل لكل مستجيبAI needs complete profiles, not aggregated averagesالذكاء الاصطناعي يحتاج ملفات كاملة، لا متوسطات مجمّعة
Segmentation question clearly labeledسؤال التجزئة محدد بوضوحTells AI which variable to use as primary clustering axisيخبر الذكاء الاصطناعي بالمتغيّر الذي يستخدمه كمحور تجميع رئيسي
Your persona framework (Wheeler elements)إطار شخصيتك (عناصر Wheeler)Gives AI the output structure you needيمنح الذكاء الاصطناعي هيكل المخرجات الذي تحتاجه
Any existing persona hypothesesأي فرضيات شخصية موجودةAI will confirm, refine, or contradict them with evidenceالذكاء الاصطناعي سيؤكدها أو يُدقّقها أو يتعارض معها بأدلة

What AI returns ما يعيده الذكاء الاصطناعي

Output elementعنصر المخرج Exampleمثال
Cluster sizeحجم التجمّع34%
Defining characteristicsالخصائص المميزةSenior, time-poor, high AI curiosity, low AI confidenceكبير المستوى، ضيق الوقت، فضول عالٍ، ثقة منخفضة بالذكاء الاصطناعي
Primary goalالهدف الأساسيEfficiency without losing creative controlالكفاءة دون فقدان السيطرة الإبداعية
Dominant pain pointنقطة الألم السائدةDoesn't trust AI output enough to publish without heavy editingلا يثق بمخرجات الذكاء الاصطناعي بما يكفي للنشر دون تحرير مكثف
Preferred channelsالقنوات المفضلةLinkedIn, email newsletters, podcasts during commuteلينكدإن، النشرات البريدية، البودكاست أثناء التنقل
Voice sampleعينة الصوتDirect quotes from open-ended answersاقتباسات مباشرة من الإجابات المفتوحة
Content implicationانعكاس المحتوىNeeds proof-based content — case studies, before/afterيحتاج محتوى قائمًا على الأدلة — دراسات حالة، قبل/بعد

Behavioral pattern detection اكتشاف الأنماط السلوكية

  • Respondents who use Tool A also consistently report Pain Point B.المستجيبون الذين يستخدمون الأداة A يُبلّغون باستمرار عن نقطة الألم B.
  • Respondents in Segment X make decisions differently than Segment Y, even when they share the same goal.المستجيبون في الشريحة X يتخذون قرارات بشكل مختلف عن الشريحة Y، حتى عندما يشتركون في نفس الهدف.
  • A specific combination of demographics + behavior predicts a specific channel preference.مزيج محدد من الديموغرافيا والسلوك يتنبأ بتفضيل قناة محدد.
These are connections a human analyst would miss in a 300-row spreadsheet. هذه روابط سيفوتها المحلل البشري في جدول بيانات من 300 صف.

The living persona الشخصية الحية

Update triggerمُحفّز التحديث AI actionإجراء الذكاء الاصطناعي
New survey wave completedاكتمال موجة استبيان جديدةCompare new clusters to previous — flag shiftsمقارنة التجمّعات الجديدة بالسابقة — تحديد التحوّلات
Focus group transcripts addedإضافة نصوص مجموعة التركيزEnrich persona voice and emotional depthإثراء صوت الشخصية والعمق العاطفي
Social listening data fed inإدخال بيانات الاستماع الاجتماعيUpdate channel preferences and language patternsتحديث تفضيلات القناة وأنماط اللغة
Customer journey data addedإضافة بيانات رحلة العميلAdd behavioral reality layer to survey-based assumptionsإضافة طبقة واقع سلوكي للافتراضات القائمة على الاستبيان
A persona built once and never updated is a liability, not an asset. الشخصية المبنية مرة واحدة ولا تُحدَّث أبدًا هي عبء لا أصل.
Case study · Unit دراسة حالة · وحدة II.15

End-to-end with AI — a startup mini-market in a residential tower من الفكرة إلى لوحة المعلومات — سوبر ماركت ناشئ داخل برج سكني

A short walk-through of the full loop we covered in this block: prompt the AI to draft a survey, collect responses, then prompt the AI again to turn the responses into a dashboard the owner can actually use. جولة قصيرة في الحلقة الكاملة التي غطّيناها في هذا المحور: نطلب من الذكاء الاصطناعي صياغة استبيان، نجمع الردود، ثم نطلب منه مجدداً تحويل الردود إلى لوحة معلومات يستطيع المالك استخدامها فعلاً.

Step 1 — Prompt used to draft the survey الخطوة 1 — الموجِّه المستخدم لصياغة الاستبيان

You are helping a startup mini-market inside a residential
tower design a customer survey. The owner wants to use the
responses to (a) identify customer personas and (b) understand
their key needs.

Draft a short survey (under 5 minutes to complete) with a mix
of multiple-choice and a couple of open-ended questions.
Group questions into clear sections.

Before you draft, ask me any questions that would meaningfully
change the survey.
أنت تساعد سوبر ماركت صغير ناشئ داخل برج سكني في تصميم استبيان للعملاء.
يرغب المالك في استخدام الردود من أجل (أ) تحديد شخصيات العملاء و(ب) فهم
احتياجاتهم الرئيسية.

قم بصياغة استبيان قصير (يستغرق إكماله أقل من ٥ دقائق) يتضمن مزيجاً من
أسئلة الاختيار من متعدد وعدد قليل من الأسئلة المفتوحة. قم بتجميع الأسئلة
في أقسام واضحة.

قبل أن تبدأ الصياغة، اطرح عليّ أي أسئلة من شأنها أن تؤثر بشكل جوهري
على الاستبيان.

Step 2 — The generated survey الخطوة 2 — الاستبيان الناتج

The model returned a clean, sectioned survey that was ready to send. Open it in a new tab and try it as if you were a resident of the tower. أعاد النموذج استبياناً نظيفاً مُقسَّماً إلى أقسام وجاهزاً للإرسال. افتحه في علامة تبويب جديدة وجرّبه كأنك أحد سكان البرج.

Open the survey ↗ افتح الاستبيان ↗ Open the survey results ↗ افتح نتائج الاستبيان ↗

Step 3 — Prompt used to analyze the responses الخطوة 3 — الموجِّه المستخدم لتحليل الردود

# Context
A survey was run for a startup mini-market inside a residential
tower. Its purpose was to help the store owner identify customer
personas and understand the key needs of potential buyers.

# Mission
Using the attached responses file, build an interactive dashboard
for me (the store owner) that surfaces insights into the potential
buyer personas and their motivations. The dashboard should also
give me access to the full raw responses for deeper exploration.

Before you start, ask me any questions that would meaningfully
change the dashboard — for example, about the file structure,
which personas I already suspect exist, or which decisions I'm
trying to make from the results.
# السياق
تم إجراء استبيان لسوبر ماركت صغير ناشئ داخل برج سكني. كان الهدف منه
مساعدة مالك المتجر على تحديد شخصيات العملاء وفهم الاحتياجات الرئيسية
للمشترين المحتملين.

# المهمة
باستخدام ملف الردود المرفق، قم ببناء لوحة معلومات تفاعلية لي (مالك
المتجر) تُبرز رؤى حول شخصيات المشترين المحتملين ودوافعهم. يجب أن
تتيح لي لوحة المعلومات أيضاً الوصول إلى الردود الكاملة لاستكشاف أعمق.

قبل أن تبدأ، اطرح عليّ أي أسئلة من شأنها أن تؤثر بشكل جوهري على لوحة
المعلومات — على سبيل المثال، حول بنية الملف، أو الشخصيات التي أتوقع
وجودها بالفعل، أو القرارات التي أحاول اتخاذها بناءً على النتائج.

Step 4 — The resulting dashboard الخطوة 4 — لوحة المعلومات الناتجة

The model returned an interactive dashboard with persona clusters, headline metrics, and a drill-through to the raw responses. This is the deliverable the owner uses to decide what to stock and who to talk to. أعاد النموذج لوحة معلومات تفاعلية تتضمن تجمّعات الشخصيات، مؤشرات رئيسية، وإمكانية التعمّق في الردود الخام. هذه هي المخرجات التي يستخدمها المالك ليقرر ماذا يخزّن ومع مَن يتحدث.

Open the persona discovery dashboard ↗ افتح لوحة اكتشاف الشخصيات ↗

Step 5 — Going deeper: analysing one open-ended question (Q14) الخطوة 5 — التعمّق: تحليل سؤال مفتوح واحد (السؤال 14)

Open-ended responses used to be expensive to read at scale — this is exactly where AI's classification and synthesis change the economics. Below is the prompt the owner used to focus the model on a single open-ended question and surface the themes inside it, including the long-tail signals a quick read would miss. كانت قراءة الردود المفتوحة على نطاق واسع مكلفة في السابق — وهنا تحديداً يُغيّر الذكاء الاصطناعي الاقتصاد عبر التصنيف والتوليف. أدناه الموجِّه الذي استخدمه المالك لتركيز النموذج على سؤال مفتوح واحد واستخراج المواضيع داخله، بما فيها الإشارات الهامشية التي قد تفوت القراءة السريعة.

# Context
A startup supermarket inside a residential tower ran a survey
to better understand its potential customers. I want to focus
exclusively on Question 14: "Imagine the store is open and
you're walking in. What's the ONE thing you'd want to find
that would make you a regular?"

# Mission
Build a dashboard that analyses the responses to Question 14
only. It should:

  - Group and categorise the responses into clear themes,
    with frequency counts for each theme.
  - Surface standout insights, including ones mentioned by
    only a few respondents if they reveal something the owner
    would otherwise miss.
  - Highlight any other interesting patterns — for example,
    recurring language, unexpected combinations, or signals
    that point to specific customer types.
  - Give me access to the underlying responses within each
    theme so I can read the raw quotes.

# Constraints
Focus only on Question 14. Ignore all other questions in the file.

Before you start, ask me any questions that would meaningfully
change the dashboard — for example, about the file structure,
the language(s) of the responses, or the decisions I'm trying
to make from these answers.
# السياق
أجرى سوبر ماركت ناشئ داخل برج سكني استبياناً لفهم عملائه المحتملين بشكل
أفضل. أرغب في التركيز حصرياً على السؤال رقم 14: "تخيّل أن المتجر مفتوح
وأنك تدخل إليه. ما هو الشيء الوحيد الذي تودّ أن تجده ويجعلك زبوناً دائماً؟"

# المهمة
قم ببناء لوحة معلومات تحلّل ردود السؤال 14 فقط. يجب أن:

  - تجمّع الردود وتصنّفها ضمن مواضيع واضحة، مع ذكر عدد التكرارات لكل
    موضوع.
  - تُبرز الرؤى اللافتة، بما في ذلك تلك التي ذكرها عدد قليل من المشاركين
    إذا كانت تكشف شيئاً قد يفوت المالك ملاحظته.
  - تُسلّط الضوء على أي أنماط مثيرة للاهتمام — مثل العبارات المتكررة،
    أو التركيبات غير المتوقعة، أو الإشارات التي تدل على أنواع معيّنة من
    العملاء.
  - تتيح لي الوصول إلى الردود الأصلية ضمن كل موضوع حتى أتمكّن من قراءة
    الاقتباسات الخام.

# القيود
ركّز على السؤال 14 فقط. تجاهل جميع الأسئلة الأخرى في الملف.

قبل أن تبدأ، اطرح عليّ أي أسئلة من شأنها أن تؤثر بشكل جوهري على لوحة
المعلومات — على سبيل المثال، حول بنية الملف، أو لغة (لغات) الردود،
أو القرارات التي أحاول اتخاذها بناءً على هذه الإجابات.

Notice the shape of the prompt: a tight scope (one question only), explicit asks for both frequency-ranked themes and long-tail signals, plus a requirement to keep the raw quotes accessible so the owner can verify any theme by reading the underlying responses. The same pattern works for any single open-ended question in any survey. لاحظ شكل الموجِّه: نطاق ضيّق (سؤال واحد فقط)، طلبات صريحة لكلٍّ من المواضيع المُرتَّبة بالتكرار والإشارات الهامشية، إضافةً إلى اشتراط إبقاء الاقتباسات الخام في المتناول حتى يستطيع المالك التحقّق من أي موضوع بقراءة الردود الأصلية. نفس النمط يعمل مع أي سؤال مفتوح واحد في أي استبيان.

The point: the owner asked AI to design the survey, write the analysis approach, and build the dashboard — three jobs that previously needed three different professionals. الفكرة: طلب المالك من الذكاء الاصطناعي تصميم الاستبيان، وكتابة منهجية التحليل، وبناء لوحة المعلومات — ثلاث مهام كانت تتطلب سابقاً ثلاثة مهنيين مختلفين.
Delivered by تُقدّمها

Galal & Karawi Management Consulting — a Dubai AI Seal-certified firm. جلال وكراوي للاستشارات الإدارية — شركة معتمدة بختم دبي للذكاء الاصطناعي.

This workshop is delivered by Galal & Karawi Management Consulting, DMCC, a firm verified under the Dubai AI Seal — the Emirate's official certification of responsible AI use across the advisory practice. Verified AI Supplier and Verified AI Business User, Tier E. تُقدّم هذه الورشة من قِبَل جلال وكراوي للاستشارات الإدارية، وهي شركة معتمدة بختم دبي للذكاء الاصطناعي — التحقق الرسمي للإمارة لاستخدام الذكاء الاصطناعي المسؤول عبر الممارسة الاستشارية. مورّد ذكاء اصطناعي معتمد ومستخدم أعمال للذكاء الاصطناعي معتمد، الفئة E.

Verify the seal at dub.ai/ai-seal → تحقّق من الختم على dub.ai/ai-seal →
Dubai AI Seal · Tier E
For attendeesللمشاركين

Keep this page open during the session. أبقِ هذه الصفحة مفتوحة خلال الجلسة.

Every section collapses and expands on demand. Switch language at any time — both versions are ready. Use the anchors in the header to jump straight to a topic, and revisit the page after the session as a working reference. يمكن طيّ كل قسم وفتحه عند الحاجة. بدّل اللغة في أي وقت — النسختان جاهزتان. استخدم روابط الترويسة للانتقال المباشر إلى أي موضوع، وارجع إلى الصفحة بعد الجلسة كمرجع عملي.