dig.ai

A luxury unboxing hits 1.4 million views on TikTok. The caption is "haul from the mall." The brand is on the label in the frame for 21 of the 47 seconds. Your text-based mention tracker records zero brand mentions on that post. The true reach is 1.4 million. The measured reach is zero. When you think about it, that gap is where most brand-mention volume in your category is quietly living right now, and honestly, it's the gap most brand teams have been slowly getting used to for two years without quite naming what happened. What happened is that the medium changed, the measurement stack didn't, and the number your dashboard reports as brand-mention volume is smaller than the number of brand mentions you actually have. The best social video content impact measurement platforms fix that gap by measuring the video itself, not the wrapper. Frame-level object and logo recognition, speech-to-text on the audio, on-screen text recognition, and actor identity, combined into a true-reach number that includes every appearance of the brand. For most categories in 2026, that number is between three and eight times larger than the captioned-mention number your tool is showing you. Full breakdown for brand and insights teams: https://lnkd.in/dx9PkvVb #socialvideointelligence #brandmarketing #brandmeasurement #socialmediameasurement

Somewhere in your category right now, a creator is halfway through a 47-second TikTok about your product, and the brand name appears exactly nowhere in the caption. It's on the label in frame at second 12, spoken aloud at second 23, on the packaging at second 34, and the comments are agreeing with an opinion the creator formed on camera in a way you'd want to know about. Your dashboard is calm. The video is at 380,000 views and climbing. When you think about it, this is the shape of almost every meaningful brand signal on social in 2026, and honestly, it's the shape most brand teams have been slowly adjusting to for two years without quite naming what happened. What happened is that the medium changed, the tooling didn't, and the gap between "we have social monitoring" and "we actually see what's happening to our brand" has been widening quietly the whole time. Social Video Intelligence is the category that closes it. Frame-level detection, multimodal sentiment across audio and visual context, and evidence that traces back to the source clip. #socialvideointelligence #brandmarketing #videofirst #socialintelligence

The deepfake of the CEO started circulating just after 6am, and by the time the comms team had finished their first coffee it was already stitched into forty different reaction videos with a combined reach past a million views. When you think about it, this is the operational reality every CCO has walked into over the last two years. The threats are faster, the medium is video, the actors are coordinated, and the tools most comms teams inherited were built for a world where the only thing to watch was a Twitter mention. A decade ago, cyber risk moved from IT hygiene to the board room, and the CISO seat matured into something that looks a lot like modern incident response. The same shift is happening now to communications, and honestly, most of the CCOs we talk to have known it for a year without quite having the language for it. The framework that works is the one CISOs already run. Detect, investigate, contain, respond, review, adapted to the video-first surface where narrative attacks actually live. #corporatecomms #narrativeintelligence #executiveprotection #ccorole

A creator on TikTok unboxes a luxury bag, walks you through the stitching and the hardware, mentions a price that's a fraction of retail, and ends with the line you already know is coming, "link in bio." The bag is counterfeit, the link routes to a private messaging app, and by the time anyone reports the post the same seller has forty parallel accounts running the same script. When you think about it, this is the part of brand protection that text-based monitoring was never built for. Counterfeit sellers worked out a long time ago that social video is the cheapest customer-acquisition channel they're ever going to get, and the protections most brand teams have in place don't extend into the frame. The brand name never appears in the caption, the hashtags use "dupe" or "inspired by," and the trademark only shows up spoken aloud or shown on the dial, which means Brandwatch, Sprout, Meltwater, and Talkwalker don't surface a single one of these videos because there's nothing in the text layer to catch. Frame-by-frame scanning, logo recognition, OCR, plus speech-to-text on the sale routing language, is the detection capability that finally closes the gap. #brandprotection #counterfeitdetection #socialvideointelligence

Every intelligence team can tell you what they are watching. 👀 Very few can tell you what they are missing. 🙈 The queries you wrote, the keywords you track, the platforms you onboarded: that is the known part, and it is the smaller part. Underneath it sits everything nobody thought to look for. The narrative forming in a language you do not monitor. The story hardening inside video comments, where your tools simply cannot follow. That is the unknown, and both the biggest risks and the best opportunities tend to live there. dig solves that. On Sep 1 to 3 we will be at ISS World Asia, Singapore. If you are there, come challenge us! Bring your hardest question, the one you are afraid to know the answer to, and we will run it live at the booth. Lihi Kraus Asaf Cohen (Pizzer) Yuval Regev Omer Lev

A consumer in a focus group says she'd absolutely pay more for sustainable packaging. Six hours later, she's on TikTok showing off the cheapest dupe she could find. Both reactions are real. Only one of them is what she'd ever tell a research vendor. This is the gap social video intelligence fills, and it's not an argument against surveys or focus groups. Both are excellent at what they do. They're also, by design, asking consumers who know they're being asked, which is the part that quietly limits what you can learn from the answers. The Hawthorne effect doesn't go away because the questionnaire is well-written. Social desirability bias doesn't disappear because the moderator is good. Social video sits outside the frame entirely. Nobody filming a haul, an unboxing, or a "why I'm switching" rant is doing it because a researcher asked. The behavior is what's actually happening, and the format is the one consumers chose. #consumerinsights #socialvideointelligence #marketresearch

How to gain 84K new followers in one day. Before the World Cup quarter-final, Norwegian Air dared British Airways: loser swaps its logo for a day. BA replied, "Don't make bets you can't win." Norway lost 2-1 to England. Norwegian didn't go quiet. It switched to the BA logo, posted "It's coming home," and congratulated England. Look at that setup as a marketer. British Airways has 1.6M followers. So the "loser" spent a day broadcasting itself to a rival's audience 25 times its size, wrapped in a story the entire internet wanted to forward. None of that could have lived in a content calendar. You can't pre-produce the payoff to a bet you might lose. You can only be ready while the moment is still warm. Reading the room while it's still warm is the real skill now. That layer is what dig reads. Pressing post is still yours. The Norway football team lost. The Norwegian Air marketing team won big time.

What happens when a government controls the media, but not social video? That was the question behind yesterday’s Financial Times story. As Ukrainian drone attacks reached Moscow, the visible damage was only part of the story. The more interesting shift was happening on social video. Our analysis found that 30% of the highest engagement posts from Russia following the June 18 attacks mocked the country’s “best air defence in the world.”  Not the attacks themselves. Not the official response. The narrative. That distinction matters. Narratives rarely announce themselves. They emerge across thousands of videos, edits, reactions, and comments. Individually, they look like noise. Together, they reveal a change in public perception. This is exactly what video-first social Intelligence is built for. Traditional monitoring tells you what was posted. dig understands how public opinion is shifting before it became obvious. https://lnkd.in/dnD_6DF6

Truth isn’t competing with lies. It’s competing with virality. By the time a false narrative is debunked, millions may have already seen it, shared it, and believed it. That’s all it takes for perception to outrun reality, and for a fake story to take on a life of its own. Our CEO and Co-Founder, Ofer Familier, breaks down how false narratives take hold, spread across social video, and create real business risk: https://lnkd.in/dEK_SaK7 #Forbes #sociallistening #digitalinsights #brandreputation #narrativeintelligence #brandstrategy #disinformation #socialintelligence

Boolean queries were built for a text-first internet. The problem? Social isn’t text-first anymore. A viral TikTok doesn’t care about your keyword list. Neither does a stitch, parody, product demo, or creator rant that never mentions your brand by name. Keywords find words, narratives find meaning, and video changes everything. If your social listening strategy starts and ends with Boolean, you’re seeing a fraction of the story. Check out our latest blog on why Boolean queries just don't work: https://lnkd.in/eQyfYk6q #sociallistening #digitalinsights #brandreputation #narrativeintelligence #brandstrategy #socialintelligence