My Startup: tracking real world trends - thoughts? suggestions? | Hacker News Reader

My Startup: tracking real world trends - thoughts? suggestions?

edopter.com

10 pointsmatthewer19 comments

I like the idea of 'trendcasting' with a social element enought to have invested in the area repeatedly, so I was intrigued by your post. Unfortunately, I was disappointed in your implementation as it is now. I'm sure my criticisms are biased by my experiences (and what I'm actively looking to invest in), but I think you're not delivering on your promise or potential in this area. If you're happy with what you have, please ignore the ramblings of this grumpy old man. I think if you place social suggestions at the heart of the service it just becomes another social news site. You essentially end up with a different face on a Digg-style site, in a way, which opens itself up to gaming of the system and takes the focus away from actual trend identification and forecasting data.

Instead of a social news remake, you could become something that actually gleans insight from social data without overt promotion/edoption by individuals. People are looking for insights about trends, but they don't just want a data dump or a shill-fest. The "Soap Box" should be small or nonexistent in such a service. When you're talking about trend identification and forecasting, most customers want some degree of editorial analysis (Springwise, Teentracking focus groups, etc.) and/or some type of meta-analysis based on aggregate data, such as search data (Yahoo Buzz, Google Trends) and/or to have the tools to do an analysis for themselves (more advanced tools with NLP/ML features), possibly within a specific medium (forums, email, blogs, search, SMS, etc.) or verticals (retail investing, public relations, etc.) There is a huge market opportunity in finding sweet spots in these 'trendcasting' areas...and there is enough room for many profitable companies. I'm disappointed that there are relatively few companies that seem to be attacking the gaps, yet there are so many different takes on "me too" companies that draw from the trend and forecasting toolset (social news, reputation management, etc. as standalone services.) Why not go after the value-added services and niches and really make some money? How do you plan to make money? [Advertising?!?]

Your design looks professional, but it is very busy and makes it difficult to really get a feel for what it does and how it all works. I think if you want your site to work well in its current 'edopter' implementation, you should focus more on simplifying the interface.

Wow, great feedback. Thanks! Let me try and address some of your points.

"I think if you place social suggestions at the heart of the service it just becomes another social news site."

I agree. We focused on this first because having user input is important to us. We strongly want to avoid becoming a digg or reddit clone, but at the same time we want to avoid being based only on science and mining the data from other sources.

“When you're talking about trend identification and forecasting, most customers want some degree of editorial analysis (Springwise, Teentracking focus groups, etc.) and/or some type of meta-analysis based on aggregate data, such as search data (Yahoo Buzz, Google Trends)”

I totally agree. Our system balances our users opinions with data from outside sources (we call this the X-factor.) By combining the two data streams we are able to get (imho) the best of both worlds. As I said, the front end for the users took up a lot of time; however, we are working hard on refining our algorithm and adding more sources. We want to combine the data driven resources of a Nielsen with the forward thinking early adopters of a “look-look’ (http://www.look-look.com/)

“How do you plan to make money?”

We want to connect our users with businesses. Users can trade points for products and opportunities from the right sponsors. Someone who is into sports could have an opportunity to try out the newest snowboard. We are working on this part of the site, but expect it to be up with in a month. Sneak peak: http://edopter.com/showcase/

Thanks for the feedback. These are some great ideas, and will work on making things more clear on the site.

I'm afraid it just confused me. I'm kinda old, so maybe I'm just too far down the Zuckerberg curve to understand.

So, I saw a "trend", I guess, called "World Politics" with a picture of Obama bein' cool (like all other pictures of Obama...that dude's definitely got cool down), clicked it, and found...nothing? I mean there was a page there, but there was just a picture of a couple of foreign dudes (World Politics...makes sense I guess) and a random video. And then I felt an overwhelming feeling of, "I want those 30 seconds back".

So, reading the comments here, I'm beginning to grasp the concept of the site, and I now realize why I was so disappointed by the content I saw: It was generated by an idiot user, the most common creature in the known universe. So, the system needs to be better at solving the hard problem of helping idiot users create something useful.

Your site design is nice, if a bit busy.

The "browse" page is a much better overview than your front page, which I found intimidating and confusing. Browse, while it could be mistaken for "just another social news site" is actually comprehensible to me, and makes your site seem more interesting...it also seems to allow the genuinely interesting stuff to rise to the top, so I don't wind up on the nonsensical "World Politics" page that has nothing interesting to look at or read.

I think maybe this is a case where simplifying your application would be a net positive. Easier said than done, I know, as I'm sure every little feature on the site had someone campaigning for it at some point in the devel/beta cycle. But new users show up and see an explosion of confusing concepts and no clear entry point to get involved.

This is great feedback. We just redid the homepage, but I totally agree. It is not coming across like I had hoped. Thanks for your comment. Truly appreciated. Can you make any suggestions for what you would like to see on the homepage. How can we make it more clear?

OK, the single biggest problem is there's way too much happening. Do you really need SEVEN tabs for each trend? I can discern no difference between four of them (overview, the pitch, the buzz, in depth)...they all seem pretty much like the same thing. And, if you're a socially generated site, isn't it ALL talk? You should probably have no tabs and a simpler view all-around. If you do have tabs, maybe just "Media" and "Talk" with the first page remaining "Overview". All of that other stuff just seems to just be people making a few random "Yeah, my idea is awesome!" comments. If you want people to have that kind of ownership, maybe let them have personal profile pages where they can just say "Yeah, I'm awesome!" and then let the trends stand alone with just the talk and media.

Likewise for the front page..."What's Next?" and "The Wire". I can't even begin to guess what the difference between the two concepts might be. Too much going on, and I'm just confused.

But, if your users can make sense of this stuff...maybe I'm just too old. But I'm definitely very confused, and often irritated about not finding the interesting bits easily. As I mentioned, the Browse page is actually pretty cool...and made me want to click through. But once I clicked, I regretted it, as I didn't find what I wanted to find on the other side. (Partly because idiot users are not interesting, and partly because I was just overwhelmed by the number of links off to nothing...one can more safely assume that half of the tabs will lead to empty pages than that they are worth clicking.)

Quick background: we are just two guys (24,25) who left our jobs and started this site. We have been working on it for months, and would love some constructive feedback. Thanks!

I rather like the feel of the site. However, I think you should order trends by the percent of users that like them, not total number of users. Otherwise, every trend's popularity would be going up as your user count increases.

Plus trendcasting is rather trendy in itself, so your data would be somewhat skewed. It will be interesting to analyze the developments of trade compared to total market share data (easy to get for public companies) as a time series problem. Then you get into the business forecasting area and your data would be worth some serious cheddar.

Just my 2 cents.

One thing that seems problematic is having too many trends. How do you plan on motivating users to discover added but unknown trends on your website so that they might be able to see the light of day?

Good point. We have been debating if we should limit the number of trends a user can be in on. We were thinking 15 at once. What do you think?

I just read this month trends at trendwatching.com, but crowdsourced trend watching is great idea. Well until new ideas stop coming. Hope you can build good user base.

One thing: Ad some kind of feeds. I just wont remember the site (despite the nice design ), I don't want yet another login to remember, but might want to fallow trends and comment on a trend I am interested in. New trends, Popular trends, Random trends, etc.

Makes sense. I will put that on our list of things to do. Thanks

Interesting site. What's the revenue model like? Seems like a shiny wikipedia combined with a simple stock market, but still don't totally get it. I just signed up so I'll give you my full thoughts once I get a chance.

allow users to bet on trends and you got yourself an investor

Interesting idea, how does it work? by aggregating news on the internet? is there some complicated algorithm?

I am still a little confused so to what I would do with it however...?

Let me try and explain:

Users launch a trend that they think is going to be big or is something they are into (saving the environment, american apparel, ect.) We then track how many users on our site like it (and where they are from, age, and sex) plus aggregated information from the internet.

We are hoping to see how the trends connect with each other, and how users find new trends. We call this tracking influence.

We are constantly working on our algorithm adding more sources. We think this is a balance between users finding the newest thing, and existing data channels (does that make sense?)

Once you sign up, the goal is to earn points by getting in/starting a new big trend. This might help: http://edopter.com/howitworks

http://www.google.com/trends

Who did the design? Because it's good.

Also, this seems reminiscent of 43things. What sites inspired you?

My cofounder did. I will let him answer the second question.

Reply on news.ycombinator.com