There are few things that are more frustrating in life than having a problem and not feeling like you’re being heard. It’s even worse when a company sends a robot or auto-generated message your way. While Amazon says it’s trying to fix this, it’s hard to believe most of the time that any messages are being read by actual human beings.
I ran into this issue when a review of mine was taken down without any explanation.
I posted a while back about how I was reposted by a large clothing company and, to my disappointment, it resulted in much less attention than I expected. This experience has been repeated on multiple occasions, but it’s also been challenged.
Earlier this year I responded to a girl’s story on Instagram (it was about guys being jerks to her about being a programmer) with the words “Thank you” which prompted her to look at my profile and message me. She’s a fellow software developer and has a rather large following on Instagram. For whatever reason she liked my profile, screenshotted it, and shared it with her followers
I was a little in awe that she responded to my innocuous message in the first place, so to see her post about my profile completely floored me. That’s also when I began getting bombarded with followers who were obviously checking out my profile from her link. For the next 24 hours every time I opened up Instagram I had a ridiculous number of notifications.
After the Instagram story timed out (after 24 hours) my daily follower average returned back to normal and I went on my way with about 500 more engaged followers.
What made this different than the first big repost?
I have a theory that there are two kinds of accounts:
There are accounts you follow because you like the photos (the aesthetic, the photography, the brand, etc.)
And there are accounts you follow because you like the stories (smaller businesses and people)
People feel connected to the accounts that make it feel “personal” – whether it’s because you have the same job or like the same things. When a “personal” account does a repost or gives a shoutout, it’s more like a personal introduction to a friend: “Hi, this person is cool, you should check them out”
If a brand/aesthetics-based account gives a shoutout their followers may or may not respond to it based on if they like that specific photo, product, or idea. The followers may not feel like they “need” to follow someone else if it’s the same aesthetic as the brand because they’re already following the brand.
After I reached a reviewer ranking of less than 80,000 I started getting emails from Amazon sellers asking me to review their products. From an algorithmic standpoint this makes complete sense: products with more (and better) reviews are more likely to be seen by other customers and the best way to advertise to potential customers is by utilizing a system already in place on the website. Sellers will offer cash or heavy discounts to entice people to review their products – some ask for “good” reviews, others just ask for reviews.
Awkwardly, this is completely against Amazon’s stated policy, but it wasn’t always. Continue reading →
If you have an internet connection, there’s a pretty good chance that at some point in time you’ve ordered something from Amazon. Amazon is great for retail consumers, it’s great for sellers, it’s great for software developers (Amazon Web Services), it’s great for affiliate marketers, it’s great for people who love instant gratification.
Point being that everyone uses Amazon, but not everyone uses it purposefully. A few months ago I happened to stumble across a few articles about Amazon Top Reviewers and the perks that come with having that title (the perks mostly include getting free things to review). After some hard-hitting research (read: clicking around Amazon) I found Amazon’s ranking criteria:
My first impression: My project manager kept saying we should do Selenium tests to increase our unit test coverage on the front end. Our previous experience was with Karma and Jasmine (which was also kind of painful) so we tried to put it off as much as possible in the interest of doing “actual work”
Second take: Oh my gosh this is witchcraft. When you use Selenium you can basically program a ghost of yourself that is better and faster that can interact with your application and notices when things break (as long as you tell it what to notice).
After hitting 1000 followers I didn’t feel any spectacular change or get any life changing epiphanies, but around 1,100 followers I slowly started to notice a confusing pattern emerge: the photos that I was posting weren’t getting the same engagement that they had been getting in the past. They were the same photos I’d always posted at the same times, but a pet photo that could have gotten seventeen ‘likes’ in four minutes was suddenly taking ten or fifteen minutes to reach that same level. At first I thought maybe there was a delay in the hashtags displaying, but I noticed the photos were still getting a significant number of views so it probably wasn’t an exposure problem.
This morning I woke up to a message from an account I follow on Instagram:
Hi Johna! I just read your bio, one of our baristas was born and raised in Alaska also! What a small world. We just had a blogger mixer and are starting to plan for our next one for April. Let me know if you’d be interested in attending.
Since my interactions with this coffee shop have been purely based on me liking their photos, I’m incredibly excited that they reached out. It just goes to show that companies do look at people’s profiles (and they will reach out).
I’m also low-key excited that they think I’m a blogger – hopefully I can attend the event and I’ll get to write a post about what that is like.
Note: There’s no way that this post will be all inclusive or up to date because of the huge scope of the internet and new websites being made every day. Feel free to reach out to me with your favorite websites/strategies if you have them.
Posting photos and being able to see likes and comments is great, and after doing it for a while you’ll be able to have a mental gist of what works and what doesn’t. Having a gist is great, but it doesn’t allow you to quantify the subtle differences (ex: I might know that pictures relating to programming do well, but maybe pictures of computers in coffee shops do better than computers on desks).
I didn’t find this out until this month, but Instagram has a lot of analytics it doesn’t make public in it’s API (this is where other websites are drawing their analytics from). For instance, the Instagram API doesn’t actually share gender of user or how many views a photo gets – if a website offers that then they’re probably trying to sort gender by something like the bio or picture recognition. Continue reading →