|
| 1 | +# Cookbook Editorial Rulebook |
| 2 | + |
| 3 | +The rules for what goes into the Data Engineering Cookbook, what stays out, and |
| 4 | +how everything cross-links. Read this before adding or restructuring content. |
| 5 | +If an edit breaks a rule here, the rule wins — or we change the rule on purpose. |
| 6 | + |
| 7 | +--- |
| 8 | + |
| 9 | +## 1. What the Cookbook IS |
| 10 | + |
| 11 | +The Cookbook is the **wide map of the field** and the **on-ramp**. Its job is to |
| 12 | +help a reader understand the 2026 data-engineering landscape, locate themselves |
| 13 | +in it, and learn *how to think* — then route them to the right deeper resource. |
| 14 | + |
| 15 | +It is the **discovery layer**: the broad, browsable, GitHub/SEO surface that |
| 16 | +pulls strangers in and turns them into email subscribers. |
| 17 | + |
| 18 | +The Cookbook owns: |
| 19 | +- **The 2026 worldview** — how the field splits (data / analytics / ML / AI |
| 20 | + engineering), what changed, where a reader fits. |
| 21 | +- **Roadmaps / "Start Here"** — the on-ramp paths by role and level. |
| 22 | +- **Judgment & decisions** — "lake vs warehouse vs lakehouse," "stream vs |
| 23 | + batch," "do you even need big data." The *why* and *how to choose*. |
| 24 | +- **Patterns & concepts** — mental models, architecture thinking, trade-offs. |
| 25 | +- **The breadth net** — case studies, interview questions, data sources. The |
| 26 | + surface that attracts traffic. |
| 27 | + |
| 28 | +## 2. What the Cookbook is NOT |
| 29 | + |
| 30 | +- **NOT a tool catalog.** It never lists "all the tools for the Process phase." |
| 31 | + That is the Tools Guide's job (see §3). When a reader asks "but which tool?", |
| 32 | + the Cookbook *links to the Tools Guide* — it does not answer inline. |
| 33 | +- **NOT a build manual.** Step-by-step pipeline builds live in courses / |
| 34 | + hands-on content, not here. |
| 35 | +- **NOT a dumping ground for links.** Resources are woven in at the point of |
| 36 | + curiosity (see §5), not collected in an end chapter. |
| 37 | + |
| 38 | +> **The one-sentence test:** if a paragraph could be replaced by a row in the |
| 39 | +> Tools Guide matrix, it does not belong in the Cookbook. |
| 40 | +
|
| 41 | +--- |
| 42 | + |
| 43 | +## 3. Division of labor (Cookbook vs Tools Guide vs the rest) |
| 44 | + |
| 45 | +| Layer | Document | Owns | Funnel role | |
| 46 | +|---|---|---|---| |
| 47 | +| **Map** | **Cookbook** | Worldview, roadmaps, decisions, concepts, breadth net | Discovery → email capture | |
| 48 | +| **Mechanics** | **Tools Guide** | Blueprint phases, tool matrix, pipeline patterns, project→course links | Focused magnet → Academy | |
| 49 | +| **Depth** | **Free content** (YT / LinkedIn / Udemy / free courses) | How-to, demos, tutorials | Woven inline everywhere | |
| 50 | +| **Paid apex** | **Academy → Masterclass** | Full courses, certification | Apex CTA | |
| 51 | + |
| 52 | +**Rule:** each fact has exactly one home. The blueprint *mechanics and tool |
| 53 | +catalog* live in the Tools Guide. The Cookbook references them; it never |
| 54 | +re-creates them. |
| 55 | + |
| 56 | +When the Cookbook touches the blueprint (Connect → Buffer → Process → Store → |
| 57 | +Visualize), it stays at the **decision/concept** altitude and hands off: |
| 58 | + |
| 59 | +> *"…for the full tool catalog and cloud-native vs open-source alternatives for |
| 60 | +> this phase, grab the Tools Guide → {{TOOLS_GUIDE_CTA}}"* |
| 61 | +
|
| 62 | +That hand-off is also a funnel step (email capture / Academy jump). One move, |
| 63 | +both goals. |
| 64 | + |
| 65 | +--- |
| 66 | + |
| 67 | +## 4. The funnel ladder (where every CTA points) |
| 68 | + |
| 69 | +Every piece of content should make the next step obvious. The ladder: |
| 70 | + |
| 71 | +1. **Cookbook** (free, broad) → capture the **email via download**. |
| 72 | +2. **Tools Guide** (focused magnet) → another capture / Academy jump. |
| 73 | +3. **Free content** (YT / LinkedIn / courses) → woven inline at each decision. |
| 74 | +4. **Academy → Masterclass** (paid apex) → the deep "go further" CTA. |
| 75 | + |
| 76 | +Goals, in order: build trust → capture email → drive views to (sponsored) free |
| 77 | +content → upsell to the paid tier. |
| 78 | + |
| 79 | +--- |
| 80 | + |
| 81 | +## 5. The inline-link pattern (how resources are woven in) |
| 82 | + |
| 83 | +Resources appear **at the point of curiosity**, never in a separate chapter. |
| 84 | +Model: the Tools Guide's example-projects move — concept → diagram → "here's the |
| 85 | +course." |
| 86 | + |
| 87 | +After a concept or decision, attach the relevant exits: |
| 88 | + |
| 89 | +```markdown |
| 90 | +## Storing your data: lake, warehouse, or lakehouse? |
| 91 | + |
| 92 | +[2–3 paragraphs of opinionated judgment — when each makes sense, the trade-off |
| 93 | +that trips people up, what I'd pick and why.] |
| 94 | + |
| 95 | +> 🎥 Watch: [video] |
| 96 | +> 📝 Read: [article / LinkedIn post] |
| 97 | +> 🧰 Full tool catalog for this phase: {{TOOLS_GUIDE_CTA}} |
| 98 | +> 🎓 Build it hands-on: [free course] |
| 99 | +> 🚀 Go deeper: {{MASTERCLASS_CTA}} |
| 100 | +``` |
| 101 | + |
| 102 | +Rules for inline links: |
| 103 | +- Only attach links that genuinely fit the topic. No padding. |
| 104 | +- A topic with no link yet = a visible gap = your content backlog. Leave the |
| 105 | + block; it tells you what to record next. |
| 106 | +- Prose carries the value. Links are exits, not the substance. |
| 107 | + |
| 108 | +--- |
| 109 | + |
| 110 | +## 6. CTA tokens (define once, reuse everywhere) |
| 111 | + |
| 112 | +Never hard-code campaign links inline. Use tokens so a single edit re-points the |
| 113 | +whole book — critical for the planned Academy→free / Masterclass→paid switch. |
| 114 | + |
| 115 | +| Token | Means | Today points to | |
| 116 | +|---|---|---| |
| 117 | +| `{{DOWNLOAD_CTA}}` | Get the PDF / join the mailing list | email-capture page | |
| 118 | +| `{{TOOLS_GUIDE_CTA}}` | Get the Tools Guide | Tools Guide download | |
| 119 | +| `{{MASTERCLASS_CTA}}` | The paid apex | Academy (later: Masterclass) | |
| 120 | +| `{{ACADEMY_CTA}}` | The course platform | LearnDataEngineering.com | |
| 121 | + |
| 122 | +When the business model changes, update the token definitions in one place — not |
| 123 | +the content. |
| 124 | + |
| 125 | +--- |
| 126 | + |
| 127 | +## 7. Modernization rules (keeping it "2026") |
| 128 | + |
| 129 | +- **Lead with current.** Modern Data Stack, lakehouse + table formats |
| 130 | + (Iceberg / Delta / Hudi), the AI-data layer (vector DBs, embeddings, RAG, |
| 131 | + feature stores), DataOps / observability / contracts / catalogs. |
| 132 | +- **Don't delete history — relabel it.** Legacy tech (MapReduce, Sqoop, Flume, |
| 133 | + Storm, Samza, Drill, Solr) becomes a short *"Legacy / know-it-exists"* note, |
| 134 | + not a deep section. |
| 135 | +- **Reframe "Data Science" → the four destination roles** (data / analytics / |
| 136 | + ML / AI engineering) wherever the old framing appears. See §9 for how these |
| 137 | + relate to the origin-based roadmaps — don't conflate the two. |
| 138 | +- **Surface, don't bury** (see §10). Hot, in-demand topics — AI engineering |
| 139 | + above all — get direct entry points near the top, not a subsection three |
| 140 | + levels deep. |
| 141 | +- **Opinion over completeness.** A clear recommendation beats an exhaustive |
| 142 | + survey. Surveys are the Tools Guide's job. |
| 143 | + |
| 144 | +--- |
| 145 | + |
| 146 | +## 8. Quick decision guide ("where does this go?") |
| 147 | + |
| 148 | +- Explaining *why* or *how to choose* → **Cookbook**. |
| 149 | +- Listing *which tools exist* for a phase → **Tools Guide** (link to it). |
| 150 | +- Step-by-step *how to build* it → **course / free content** (link to it). |
| 151 | +- A real resource (video / article / course) → **inline, at the topic**. |
| 152 | +- "All the resources in one place" → **no.** See §5. |
| 153 | + |
| 154 | +--- |
| 155 | + |
| 156 | +## 9. Paths & roadmaps: two axes (don't conflate them) |
| 157 | + |
| 158 | +Two different questions, two different structures. Mixing them is the most |
| 159 | +likely structural mistake — keep them separate. |
| 160 | + |
| 161 | +| Axis | Answers | Keyed on | Owner | |
| 162 | +|---|---|---|---| |
| 163 | +| **Destination** | "What are the DE-adjacent roles in 2026, which do I want?" | where you're **going** | Cookbook-owned field map | |
| 164 | +| **Origin** | "I'm an analyst / dev / grad — how do I get in from here?" | where you're **coming from** | mirrors the LDE roadmaps 1:1 | |
| 165 | + |
| 166 | +**Destination → the field map.** A Cookbook-owned worldview section: data / |
| 167 | +analytics / ML / AI engineering as distinct roles. Pure orientation, no funnel |
| 168 | +pressure — it serves the reader who doesn't yet know what they want. No LDE |
| 169 | +dependency. |
| 170 | + |
| 171 | +**Origin → the "Start Here" roadmaps.** These mirror the LDE roadmap selector |
| 172 | +1:1 (Beginner / Data Analyst / Data Scientist / Software Developer). The Cookbook |
| 173 | +gives the *free orientation* for each ("the path from where you are"); each one's |
| 174 | +apex CTA deep-links to the matching **paid** LDE roadmap. The Cookbook never |
| 175 | +reproduces the week-by-week curriculum (§2). |
| 176 | + |
| 177 | +- Roadmap selector: https://learndataengineering.com/p/roadmap-selector |
| 178 | +- Canonical origin roadmaps = Beginner · Data Analyst · Data Scientist · |
| 179 | + Software Developer. Name them to match LDE so the funnel is 1:1. |
| 180 | + |
| 181 | +**AI Engineering straddles both.** On LDE it's a *destination* ("become an AI |
| 182 | +engineer"), not an origin like the other four. So it appears in the field map as |
| 183 | +a first-class role **and** as a surfaced roadmap on-ramp — flagged as "where |
| 184 | +you're going," not "where you came from." |
| 185 | + |
| 186 | +--- |
| 187 | + |
| 188 | +## 10. Surface, don't bury |
| 189 | + |
| 190 | +On a five-year-old wide map, the topics people actually search for get buried |
| 191 | +under accumulated structure unless we deliberately surface them. Don't let that |
| 192 | +happen — give in-demand topics direct entry points near the top, the same way |
| 193 | +the LDE site routes people straight to AI engineering instead of hiding it. |
| 194 | + |
| 195 | +**Worked example — AI engineering appears in three surfaced places, not one:** |
| 196 | +- In the **field map** as a first-class destination role (§9). |
| 197 | +- As a **direct on-ramp / CTA** to the AI Engineering roadmap. |
| 198 | +- As the **AI-data layer** woven through the blueprint — vector DBs, embeddings, |
| 199 | + RAG, feature stores — connecting to the existing Ollama/RAG tutorials in |
| 200 | + `sections/04-HandsOnCourse.md`. |
| 201 | + |
| 202 | +**Why it matters for the funnel:** the surfaced AI on-ramp *is* a top-of-funnel |
| 203 | +magnet, because it matches what strangers are searching for in 2026. Burying it |
| 204 | +wastes the exact traffic we most want to capture. |
| 205 | + |
| 206 | +**Rule of thumb:** if a topic is in high demand and a reader would expect to find |
| 207 | +it fast, it gets a surfaced entry point — not a deep subsection. |
0 commit comments