MoburstAI commited on
Commit
b90c03f
Β·
verified Β·
1 Parent(s): 23e3186

Creative strategy insights: translate brain regions to actionable ad advice

Browse files
Files changed (1) hide show
  1. app.py +70 -28
app.py CHANGED
@@ -212,65 +212,107 @@ def find_peak_region_seconds(region_activations, hrf_offset):
212
  return summary
213
 
214
 
215
- def generate_insights(engagement, hrf_offset=5):
 
216
  insights = []
217
  if len(engagement) < 3:
218
  return insights
219
 
220
  mean_eng = float(np.mean(engagement))
 
221
 
222
- # Hook analysis (first 3s)
223
  hook = float(np.mean(engagement[:3]))
224
  if hook < mean_eng * 0.85:
225
  drop = ((mean_eng - hook) / mean_eng * 100)
226
  insights.append({
227
- "type": "hook_weakness", "severity": "high",
228
- "message": f"Weak opening: first 3 seconds are {drop:.0f}% below the video's average engagement.",
229
- "suggestion": "Lead with your strongest visual moment or audio hook in the first 2 seconds."
 
230
  })
231
  elif hook > mean_eng * 1.15:
232
  lift = ((hook - mean_eng) / mean_eng * 100)
233
  insights.append({
234
- "type": "strong_hook", "severity": "positive",
235
- "message": f"Strong hook: opening 3 seconds drive {lift:.0f}% more engagement than average.",
236
- "suggestion": "This opening grabs attention effectively β€” keep this pattern."
 
237
  })
238
 
239
- # Mid-section drop
240
- if len(engagement) >= 8:
241
- mid_s = len(engagement) // 3
242
- mid_e = 2 * len(engagement) // 3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
243
  mid_eng = float(np.mean(engagement[mid_s:mid_e]))
244
  if mid_eng < mean_eng * 0.8:
245
  insights.append({
246
  "type": "mid_drop", "severity": "medium",
247
- "message": f"Engagement dips in the middle section (seconds {mid_s+hrf_offset}-{mid_e+hrf_offset}).",
248
- "suggestion": "Add a pattern interrupt β€” new visual, text overlay, scene change, or audio shift."
 
249
  })
250
 
251
- # Ending strength
252
- if len(engagement) >= 5:
253
  last_3s = float(np.mean(engagement[-3:]))
254
  if last_3s < mean_eng * 0.8:
255
  insights.append({
256
- "type": "weak_ending", "severity": "medium",
257
- "message": "Engagement drops off at the end where the CTA typically appears.",
258
- "suggestion": "Strengthen the closing β€” add urgency, a bold visual, or a direct CTA."
 
259
  })
260
- elif last_3s > mean_eng * 1.2:
261
  insights.append({
262
- "type": "strong_ending", "severity": "positive",
263
- "message": "Strong finish β€” engagement rises at the end, which is ideal for CTA delivery.",
264
- "suggestion": "The ending builds well. Viewers are primed for action when the CTA hits."
 
265
  })
266
 
267
- # Variability insight
268
  std_eng = float(np.std(engagement))
269
  if std_eng < 0.05:
270
  insights.append({
271
- "type": "flat_engagement", "severity": "medium",
272
- "message": "Brain response is relatively flat throughout β€” no strong peaks or valleys.",
273
- "suggestion": "The ad plays safe. Add a surprising visual moment or emotional beat to create a peak."
 
274
  })
275
 
276
  return insights
@@ -359,7 +401,7 @@ def analyze_video(video_url: str, filename: str = "video.mp4") -> str:
359
  peak_idx = int(np.argmax(valid_engagement))
360
  low_idx = int(np.argmin(valid_engagement))
361
 
362
- insights = generate_insights(valid_engagement, hrf_offset)
363
  elapsed = time.time() - start_time
364
 
365
  report = {
 
212
  return summary
213
 
214
 
215
+ def generate_insights(engagement, region_activations, hrf_offset=5):
216
+ """Generate actionable creative strategy insights from brain data."""
217
  insights = []
218
  if len(engagement) < 3:
219
  return insights
220
 
221
  mean_eng = float(np.mean(engagement))
222
+ n = len(engagement)
223
 
224
+ # ── 1. Hook Analysis ──
225
  hook = float(np.mean(engagement[:3]))
226
  if hook < mean_eng * 0.85:
227
  drop = ((mean_eng - hook) / mean_eng * 100)
228
  insights.append({
229
+ "type": "hook", "severity": "high",
230
+ "title": "Weak Opening Hook",
231
+ "message": f"The first 3 seconds generate {drop:.0f}% less brain engagement than the rest of the ad. In a feed environment, this means scroll-past risk is high.",
232
+ "action": "Front-load impact: lead with a surprising visual, bold text, or audio spike in the first 1-2 seconds."
233
  })
234
  elif hook > mean_eng * 1.15:
235
  lift = ((hook - mean_eng) / mean_eng * 100)
236
  insights.append({
237
+ "type": "hook", "severity": "positive",
238
+ "title": "Strong Opening Hook",
239
+ "message": f"The opening drives {lift:.0f}% higher engagement than average β€” this ad grabs attention immediately.",
240
+ "action": "This hook pattern works. Replicate the opening structure (visual style, pacing, audio) across other creatives."
241
  })
242
 
243
+ # ── 2. What's Driving Engagement (translate brain regions to creative language) ──
244
+ region_means = {}
245
+ for rname in ["visual_processing", "attention", "emotional_response", "language_processing", "motor_planning"]:
246
+ if rname in region_activations:
247
+ vals = region_activations[rname]
248
+ search = vals[:-2] if len(vals) > 3 else vals
249
+ region_means[rname] = float(np.mean(search))
250
+
251
+ if region_means:
252
+ sorted_regions = sorted(region_means.items(), key=lambda x: x[1], reverse=True)
253
+ top = sorted_regions[0][0]
254
+
255
+ driver_map = {
256
+ "visual_processing": ("Visually-Led Creative", "The visual elements (colors, motion, composition) generate the strongest brain response. Viewers are engaged by what they see, not what they hear.", "Double down on visual storytelling. Test different visual treatments β€” the audience is responding to imagery over copy."),
257
+ "attention": ("High Cognitive Engagement", "This ad demands active focus β€” viewers are thinking and processing, not passively watching.", "This is engaging but mentally demanding. Balance information-dense sections with breathing room to avoid fatigue."),
258
+ "emotional_response": ("Emotionally Resonant", "Emotional processing dominates β€” this ad creates a feeling. Emotion-driven ads tend to build stronger brand recall.", "Lean into the emotional angle in your CTA. 'Feel' CTAs ('Join the movement', 'Don't miss out') outperform 'think' CTAs ('Learn more') for emotional creatives."),
259
+ "language_processing": ("Copy & Audio Are Carrying", "Language processing is the dominant signal β€” voiceover, on-screen text, or dialogue are doing the heavy lifting.", "The words matter most. A/B test different copy, voiceover tones, or text overlay styles to find the strongest variant."),
260
+ "motor_planning": ("Action-Oriented", "The brain is preparing to act β€” tap, swipe, click. This is the ideal response for performance ads.", "Capitalize on this impulse. Make the CTA large, immediate, and friction-free. This audience is ready to convert."),
261
+ }
262
+
263
+ if top in driver_map:
264
+ title, msg, action = driver_map[top]
265
+ insights.append({"type": "driver", "severity": "positive", "title": title, "message": msg, "action": action})
266
+
267
+ # ── 3. Peak Moment ──
268
+ peak_idx = int(np.argmax(engagement))
269
+ peak_s = peak_idx + hrf_offset
270
+ insights.append({
271
+ "type": "peak", "severity": "positive",
272
+ "title": f"Strongest Moment at {peak_s}s",
273
+ "message": f"Peak brain engagement occurs at second {peak_s}. Whatever happens here β€” a visual reveal, scene transition, or key message β€” is the most impactful element.",
274
+ "action": f"Build your ad's narrative arc around the {peak_s}-second mark. Consider placing your product reveal or key value proposition here."
275
+ })
276
+
277
+ # ── 4. Mid-Section Drop ──
278
+ if n >= 8:
279
+ mid_s = n // 3
280
+ mid_e = 2 * n // 3
281
  mid_eng = float(np.mean(engagement[mid_s:mid_e]))
282
  if mid_eng < mean_eng * 0.8:
283
  insights.append({
284
  "type": "mid_drop", "severity": "medium",
285
+ "title": "Middle Section Loses Attention",
286
+ "message": f"Engagement drops between seconds {mid_s+hrf_offset}-{mid_e+hrf_offset}. Without new stimulus, viewers disengage here.",
287
+ "action": "Add a pattern interrupt: scene change, new speaker, text overlay, music shift, or tempo change. The brain needs novelty every 3-5 seconds."
288
  })
289
 
290
+ # ── 5. Ending / CTA Readiness ──
291
+ if n >= 5:
292
  last_3s = float(np.mean(engagement[-3:]))
293
  if last_3s < mean_eng * 0.8:
294
  insights.append({
295
+ "type": "ending", "severity": "high",
296
+ "title": "CTA Lands on Low Engagement",
297
+ "message": "Brain engagement is low when the CTA appears. Viewers' attention has drifted β€” they're less likely to act.",
298
+ "action": "Either move the CTA to the peak moment, or create a second attention spike at the end (urgency text, bold visual, or a direct question to re-engage)."
299
  })
300
+ elif last_3s > mean_eng * 1.1:
301
  insights.append({
302
+ "type": "ending", "severity": "positive",
303
+ "title": "Strong CTA Window",
304
+ "message": "Engagement is high at the end β€” viewers are attentive when the CTA appears, maximizing click-through potential.",
305
+ "action": "Ensure the CTA is visually prominent and uses action language. This audience is primed to convert."
306
  })
307
 
308
+ # ── 6. Overall Dynamics ──
309
  std_eng = float(np.std(engagement))
310
  if std_eng < 0.05:
311
  insights.append({
312
+ "type": "dynamics", "severity": "medium",
313
+ "title": "Flat Response β€” Low Memorability",
314
+ "message": "Brain activation is uniform throughout β€” no emotional peaks or valleys. Flat response correlates with lower ad recall.",
315
+ "action": "Create contrast: a quiet moment followed by a bold one, humor, an unexpected visual, or a change in pace. Memorable ads have peaks."
316
  })
317
 
318
  return insights
 
401
  peak_idx = int(np.argmax(valid_engagement))
402
  low_idx = int(np.argmin(valid_engagement))
403
 
404
+ insights = generate_insights(valid_engagement, region_activations, hrf_offset)
405
  elapsed = time.time() - start_time
406
 
407
  report = {