igerber · igerber · Jul 21, 2026 · Jul 20, 2026
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -8,6 +8,14 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ## [Unreleased]
 
 ### Added
+- **Docs: navigation and SEO polish.** A stable/latest version switcher in the navbar
+  (`docs/_static/switcher.json`); a custom `robots.txt` that keeps the ~60 thin
+  `_modules/` source-view pages out of crawlers (documentation pages allowed as
+  before); "Tutorial N:" prefixes dropped from 8 notebook H1s so page titles and
+  search-result snippets lead with the topic (page URLs unchanged - only the H1's
+  own fragment id changes, and an H1 anchor resolves to the page top that the bare
+  URL already opens; sidebar labels were already short-form); the 5 section landing
+  pages are now linked from `diff_diff/guides/llms.txt`.
 - **Docs: search-result excerpts fixed and search-as-you-type enabled.** Notebook
   section anchors containing CSS-special characters (digit-leading `#3.-Fit-Event-Study`,
   apostrophes, parentheses, colons) crashed Sphinx's search-excerpt renderer

diff --git a/diff_diff/guides/llms.txt b/diff_diff/guides/llms.txt
@@ -31,6 +31,13 @@ Full practitioner guide: call `diff_diff.get_llm_guide("practitioner")`
 
 ## Documentation
 
+The site is organized into 5 sections, each with a landing page:
+[Getting Started](https://diff-diff.readthedocs.io/en/stable/getting_started.html) ·
+[Practitioner Guide](https://diff-diff.readthedocs.io/en/stable/practitioners.html) ·
+[Tutorials](https://diff-diff.readthedocs.io/en/stable/tutorials/index.html) ·
+[User Guide](https://diff-diff.readthedocs.io/en/stable/user_guide.html) ·
+[API Reference](https://diff-diff.readthedocs.io/en/stable/api/index.html)
+
 ### Getting Started
 
 - **Practitioner Guide** (call `diff_diff.get_llm_guide("practitioner")`): 8-step workflow for rigorous DiD analysis (Baker et al. 2025) — **start here**

diff --git a/docs/_static/switcher.json b/docs/_static/switcher.json
@@ -0,0 +1,13 @@
+[
+  {
+    "name": "stable",
+    "version": "stable",
+    "url": "https://diff-diff.readthedocs.io/en/stable/",
+    "preferred": true
+  },
+  {
+    "name": "latest",
+    "version": "latest",
+    "url": "https://diff-diff.readthedocs.io/en/latest/"
+  }
+]
diff --git a/docs/conf.py b/docs/conf.py
@@ -110,6 +110,9 @@
     "../diff_diff/guides/llms-full.txt",
     "../diff_diff/guides/llms-practitioner.txt",
     "../diff_diff/guides/llms-autonomous.txt",
+    # Overrides RTD's allow-everything default at the domain root: keeps the
+    # ~60 thin _modules/ source-view pages out of crawlers.
+    "robots.txt",
 ]
 sitemap_url_scheme = "{link}"
 
@@ -129,7 +132,21 @@
             "icon": "fa-brands fa-python",
         },
     ],
-    "navbar_end": ["theme-switcher", "navbar-icon-links"],
+    "navbar_end": ["version-switcher", "theme-switcher", "navbar-icon-links"],
+    # Version dropdown (stable/latest). json_url points at the latest build
+    # so new entries propagate to every published version without rebuilds.
+    # ONE-SELECTOR POLICY: this navbar switcher replaces the RTD flyout
+    # (Settings -> Addons -> Flyout menu is disabled in the RTD dashboard,
+    # decided 2026-07-20) - re-enabling the flyout would put two version
+    # controls with different version lists on every page.
+    # check_switcher=False: the build-time URL probe would fail -W on CI and
+    # on the first RTD build (the URL only exists after this change ships);
+    # the switcher itself is fetched client-side at page load.
+    "switcher": {
+        "json_url": "https://diff-diff.readthedocs.io/en/latest/_static/switcher.json",
+        "version_match": os.environ.get("READTHEDOCS_VERSION", "latest"),
+    },
+    "check_switcher": False,
     "navigation_depth": 3,
     "show_toc_level": 2,
     # Live-filtering search overlay. Safe to enable now that search-result

diff --git a/docs/robots.txt b/docs/robots.txt
@@ -0,0 +1,10 @@
+# Served at the domain root via conf.py html_extra_path (overrides the
+# Read the Docs autogenerated robots.txt, which allows everything).
+# The ~60 autogenerated _modules/ source-view pages are thin duplicates of
+# the API reference - keep them out of crawlers so search engines surface
+# the documented pages instead.
+
+User-agent: *
+Disallow: /en/*/_modules/
+
+Sitemap: https://diff-diff.readthedocs.io/sitemap.xml
diff --git a/docs/tutorials/08_triple_diff.ipynb b/docs/tutorials/08_triple_diff.ipynb
@@ -4,7 +4,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "# Tutorial 8: Triple Difference (DDD) Estimation\n",
+    "# Triple Difference (DDD) Estimation\n",
     "\n",
     "This tutorial covers the Triple Difference (DDD) estimator, which extends standard Difference-in-Differences to settings where treatment requires satisfying two criteria.\n",
     "\n",

diff --git a/docs/tutorials/18_geo_experiments.ipynb b/docs/tutorials/18_geo_experiments.ipynb
@@ -5,7 +5,7 @@
    "id": "t18-cell-001",
    "metadata": {},
    "source": [
-    "# Tutorial 18: Geo-Experiment Analysis with SyntheticDiD\n",
+    "# Geo-Experiment Analysis with SyntheticDiD\n",
     "\n",
     "A practitioner walkthrough for marketing analytics teams measuring lift from a campaign that ran in a subset of geographic markets. The tutorial uses `SyntheticDiD` (Arkhangelsky et al. 2021) on a simulated DMA panel and walks through the fit, the diagnostics, and the stakeholder summary you need to ship the result."
    ]

diff --git a/docs/tutorials/19_dcdh_marketing_pulse.ipynb b/docs/tutorials/19_dcdh_marketing_pulse.ipynb
@@ -4,7 +4,7 @@
    "cell_type": "markdown",
    "id": "t19-cell-001",
    "metadata": {},
-   "source": "# Tutorial 19: dCDH for Marketing Pulse Campaigns\n\nA practitioner walkthrough for measuring lift from promotional campaigns that turn on AND off across markets at staggered times. The tutorial uses the `ChaisemartinDHaultfoeuille` estimator (alias `DCDH`) - diff-diff's estimator purpose-built for reversible (non-absorbing) treatment: the most general non-absorbing option in the library, with explicit joiner/leaver decomposition. (`LPDiD` and `TROP` also support non-absorbing treatment via their `non_absorbing` parameters, under stronger assumptions.)"
+   "source": "# dCDH for Marketing Pulse Campaigns\n\nA practitioner walkthrough for measuring lift from promotional campaigns that turn on AND off across markets at staggered times. The tutorial uses the `ChaisemartinDHaultfoeuille` estimator (alias `DCDH`) - diff-diff's estimator purpose-built for reversible (non-absorbing) treatment: the most general non-absorbing option in the library, with explicit joiner/leaver decomposition. (`LPDiD` and `TROP` also support non-absorbing treatment via their `non_absorbing` parameters, under stronger assumptions.)"
   },
   {
    "cell_type": "markdown",

diff --git a/docs/tutorials/20_had_brand_campaign.ipynb b/docs/tutorials/20_had_brand_campaign.ipynb
@@ -5,7 +5,7 @@
    "id": "t20-cell-001",
    "metadata": {},
    "source": [
-    "# Tutorial 20: HAD for a National Brand Campaign with Regional Spend Intensity\n",
+    "# HAD for a National Brand Campaign with Regional Spend Intensity\n",
     "\n",
     "A practitioner walkthrough for measuring per-dollar lift when every market is treated at a different dose level - treatment varies in intensity, not in status, and no never-treated unit exists. Comparison comes from the dose variation across markets, not from an untreated holdout. The tutorial uses the `HeterogeneousAdoptionDiD` estimator (alias `HAD`), built for this case."
    ]

diff --git a/docs/tutorials/21_had_pretest_workflow.ipynb b/docs/tutorials/21_had_pretest_workflow.ipynb
@@ -5,7 +5,7 @@
    "id": "d4e3e374",
    "metadata": {},
    "source": [
-    "# Tutorial 21: HAD Pre-test Workflow - Running the Pre-test Diagnostics on the Brand Campaign Panel\n",
+    "# HAD Pre-test Workflow - Running the Pre-test Diagnostics on the Brand Campaign Panel\n",
     "\n",
     "[Tutorial 20](20_had_brand_campaign.ipynb) fit `HeterogeneousAdoptionDiD` (HAD) on a regional brand-campaign panel and reported a per-dollar lift, with a brief visual placebo check at the end. We deliberately deferred the **formal pre-test workflow** to this tutorial, with a forward pointer in T20's \"Extensions\" section.\n",
     "\n",

diff --git a/docs/tutorials/22_had_survey_design.ipynb b/docs/tutorials/22_had_survey_design.ipynb
@@ -5,7 +5,7 @@
    "id": "bd4f9026",
    "metadata": {},
    "source": [
-    "# Tutorial 22: Survey-Weighted HAD - The BRFSS-Shape Rollout\n",
+    "# Survey-Weighted HAD - The BRFSS-Shape Rollout\n",
     "\n",
     "The HAD series so far ran on simple iid panels. T20\n",
     "(`docs/tutorials/20_had_brand_campaign.ipynb`) walked the headline workflow;\n",

diff --git a/docs/tutorials/24_staggered_vs_collapsed_power.ipynb b/docs/tutorials/24_staggered_vs_collapsed_power.ipynb
@@ -5,7 +5,7 @@
    "id": "269c802e",
    "metadata": {},
    "source": [
-    "# Tutorial 24: Staggered Rollout or a Simple 2×2? A Power-Analysis Decision Guide\n",
+    "# Staggered Rollout or a Simple 2×2? A Power-Analysis Decision Guide\n",
     "\n",
     "You're running a geo experiment — a feature rolled out across the **50 US states**. The\n",
     "rollout is *staggered*: a clean pre-period, then states switch on in waves over several\n",

diff --git a/docs/tutorials/27_cic_distributional_effects.ipynb b/docs/tutorials/27_cic_distributional_effects.ipynb
@@ -5,7 +5,7 @@
    "id": "f9b19c35",
    "metadata": {},
    "source": [
-    "# Tutorial 27: When the Average Hides the Action - Distributional DiD with Changes-in-Changes\n",
+    "# When the Average Hides the Action - Distributional DiD with Changes-in-Changes\n",
     "\n",
     "A loyalty program was rolled out to a lower-spend customer segment. Finance ran the\n",
     "standard difference-in-differences on monthly spend and got a flat zero - the program\n",