Posts in Artificial Intelligence

Influencer Marketing 2018 Round Up South East Asia

January 9, 2019 Posted by Artificial Intelligence, Influencer Marketing 0 thoughts on “Influencer Marketing 2018 Round Up South East Asia”

Year 2018 has been a phenomenal year for marketers in South East Asia. Over 1.5 billion active social media users have helped South East Asia catapult ahead and become one of the most sought after markets for social media influencer marketing.
We have complied influencer marketing metrics for 8 different countries in South East Asia and packaged them into a swanky report.

Download Now



Get the full report for South East Asia RIGHT NOW!

Download Now

Influencer Marketing in India

December 12, 2018 Posted by Artificial Intelligence, Influencer Marketing 0 thoughts on “Influencer Marketing in India”

Influencer Marketing in India is on the rise, and for good reason: the right influencers can make or break a brand. It is no surprise that brands in India are joining in on the trend. But will they truly get the promised ROI from their influencers?

India’s Fake Influencer Problem

At Affable, we use data to audit influencers. We have built cutting-edge machine learning algorithms to profile influencers and their audience. In India, we have profiled over 30,000 influencers (and counting!). While running through the data, we noticed that a lot of “influencers” have fake followers.

This is a problem. One of the reasons influencer marketing works so well is that influencers are authentic and resonate with their audience. However, little room for authenticity remains if the audience itself is purchased. Paying as little as ₹150 ($2.10) can earn you up to 1000 followers!

How to tell if an influencer has bought followers?

One way to figure out if an influencer has purchased followers is by sifting through their followers. If you see something like the following then there is a high chance that the followers are paid for:


Similar account names, with stock imagery as profile pictures

As you can probably see, this process is highly manual and time-consuming. Luckily, at Affable, our machine learning algorithms do that for you! This is described in our previous post.

The good news

It’s not all gloom and doom. India has plenty of authentic, engaging influencers, they’re just harder to find. Luckily for us, our advanced search engine is great at finding exactly those influencers. Here is a food influencer that our search engine identified as having one of the highest engagement levels:

Sample analytics of @the_savage_lens on available on Affable

This is influencer goes by the name @the_savage_lens on Instagram. He is a Pune based creator that takes gorgeous pictures of food. Our analysis shows that he is one of the most engaging Indian influencers, with an average engagement of 11%, which is several times higher than the 1.49% engagement of similar influencers. This, of course, is no mystery once you see how beautiful and well composed his posts are:

View this post on Instagram

This vadapav was ajajzbzbzbbzznnz….❣️ And I'm feeling ajsnsnznzn right now because a) I missed the deadlines for @platterjoy giveaway b) I'm not getting time to write a crazy good caption. So my dear dear sizzy, Nidhi, accept this entry of a Desi snack with a twist. The twist is that I made the Vada and everyone twisted after eating it because I misjudged the amount of spices. I know I'm awesome at cooking. No claps pls. 🙂 . Anywho, it's Wednesday… midweek and I've finally come home. My braveheart crossing 14 signals to go to office and another 14 to come back home. And typing this caption while waiting in between the ever growing traffic, pollution and population of Pune just to save a few minutes so that after coming home I can just post it ASAP because I gotta leave for work!🥂 . Ah! Wasn't I just back from one? Yeah, that was the one which pays my bills. This is the one which pays for my smiles. Yes, you might have guessed it right – Photography. I'm finally visiting a place I've always loved. Which place is that? Well.. keep checking the stories for details. 🐿️ . I'll roast you if you find typos 🦅

A post shared by The Savage Lens (@the_savage_lens) on

Now the only question remains, how can you find influencers as engaging as @the_savage_lens to spread the word about your brand?

Discover Authentic Indian Influencers

For an influencer marketing campaign to deliver a strong ROI, choosing the right influencer is critical. Affable’s AI-powered discovery engine gives you the ability to find the most relevant influencers for your brand. Sign up free today and search among 30,000+ Indian Instagram influencers here:

Free Trial. Easy setup. Cancel anytime.

Detect Fake Instagram Followers using AI

November 22, 2018 Posted by Artificial Intelligence, Influencer Marketing 0 thoughts on “Detect Fake Instagram Followers using AI”

Social Media platforms like Instagram and Twitter have taken a strong stance against fake followers and accounts that boost their social presence with the purchase of fake followers. Instagram’s latest announcement is another critical step towards reducing this unwelcomed behavior.

This is an excerpt from their press release – “Recently, we’ve seen accounts use third-party apps to artificially grow their audience. Every day people come to Instagram to have real experiences, including genuine interactions. It is our responsibility to ensure these experiences aren’t disrupted by inauthentic activity. Starting today, we will begin removing inauthentic likes, follows and comments from accounts that use third-party apps to boost their popularity.”

Genuine Influencers and Key Opinion Leaders are popular because of the content they create. These are content creators who amass huge following from an audience who relate to them and trust them. At Affable, we have built machine learning algorithms that can detect suspicious activity from followers and flag them. For any influencer, you can view what is the % of suspicious followers. On an average, we notice 5-15% suspicious followers for most accounts since these are just inactive accounts that end up following everyone.


Screenshot of an influencer's profile on Affable showing fake followers

How does Affable detect fake followers?

Our proprietary algorithm analyses all followers of an influencer to gauge if they have characteristics of a fake account. As humans, we can detect looking at a profile if it is suspicious – look at the profile below. The username and the staggering following-to-follower ratio clearly suggests it is a dummy account. This is not all though, our algorithm also looks at other profile criteria to gauge its authenticity.


Screenshot from Instagram with fake followers


Affable also analyzes an influencer’s follower growth over time. For any influencer, you can visualize the timeline of their follower growth. Sharp spikes typically suggest that the account has had a mass-following in a short span. Sometimes these spikes are also caused by giveaways, but you should be careful when you see concurrent spikes as below.

Screenshot of an influencer's profile on Affable showing fake followers


Brands are taking a proactive step to avoid spending marketing dollars on influencers with fake followers. Most recently, Keith Weed, Chief Marketing Officer of Unilever expressed concerns and took a stance that the Consumer Goods company will not work with any influencer who purchases ghost followers.

Are you planning an influence marketing campaign? Do you know the influencers you have in mind are real or if they have fake followers? Check them out on Affable with a free trial:

Affable : AI solution to scale word-of-mouth marketing

June 18, 2018 Posted by Artificial Intelligence 0 thoughts on “Affable : AI solution to scale word-of-mouth marketing”

For decades, Word of mouth has been the most effective mode of marketing. Traditionally, brands have been working with celebrities to narrate the story to their audiences. With the rise of popular social media users who have 10k to 100k followers, brands have a way to reach to a more targeted & niche audience at a lesser cost. These users, also called micro-influencers, are more effective than celebrities to engage with a niche audience.

Micro-influencer marketing is surprisingly effective with brands getting as much as 20% ROI. Over the last couple of years, marketers are now shifting their marketing budgets towards micro-influencers.

If influencer marketing is so effective, why isn’t everyone using it?

influencer marketing current manual process

Current process of influencer marketing


The current process of finding and working with these micro-influencers is highly manual, time-consuming and full of guesswork. Marketers spend days manually searching through hashtags finding for a relevant user, and then based on vanity metrics like number of followers & asking friends, they decide whom to engage.

Funnily, they still end up working with the wrong influencer.

Affable is an end-to-end platform that to help marketers discover, engage & measure authentic micro-influencers –

Demo link to Affable Platform

Affable Demo


At Affable, we are building computer vision algorithms to profile influencers through their images. We scan public Instagram content to gauge user-interest from their images and videos. We also analyze images uploaded by their followers to determine their demographics using advanced facial recognition algorithms.

Affable Influencer Profile Screenshot


Because we use images instead of text, the platform is language agnostic and can work in any region, speaking any language with no additional effort.

Image that shows how affable platform is language agnostic


Image that shows the engagement feature of Affable


The problem of bots and fake-accounts has plagued social media and lot of influencers seem to have a surprisingly high number of suspicious followers. We have built Machine Learning algorithms to tackle this problem — where our model can distinguish a suspicious user from a real user.

We run this analysis for all influencer’s followers to determine how many suspicious accounts follow them.



Image that shows Influencer Brand Affinity



For influencer marketing to be authentic and effective, marketers should make sure that the influencer they engage hasn’t endorsed a competing brand recently — and this happens more frequently than one would imagine.

To address this, Affable scans through all the media uploaded by an influencer to enlist the brands they have engaged with recently.



Affable’s post-campaign analytics helps marketers determine the true reach — the persona of the audience that engages with the influencer content. This helps marketers not only measure the reach of the current campaign but also helps them decide which influencers to work with, in the future.

Image that shows screen shot of post-campaign analytics on Affable


Our vision is to empower anyone — from a homepreneur to an enterprise — with tools that can scale their word-of-mouth marketing requirements. We strongly believe that technology can make it possible and at Affable we are building that technology that unlocks scale in word-of-mouth marketing.

Image that shows technology needed to scale word-of-mouth marketing

We are building our platform with a lot of love and sweat. To get an early sneak-peak into Affable, register for early access at or write to us at