A requester wants more workers (how do you think we should respond?)

[markwhiting]

The problem (or situation, in this case)

A requester who has been using Daemo to collect data has asked the Collective to add a few workers soon, to help them meet data collection goals for a deadline.

The requester thinks that they need approximately 10 more workers who each do one task (2-3hr of work) per day.

(The email from the requester can be shared in Slack)

The proposal

There's a few options I can think of, but I'd be interested to hear what others think about them, or if others have alternative suggestions. Here are the options I was thinking about:

1. Add the workers (there's a list of proactive workers we could recruit from) and solve the problem that way, however this comes with a potential downside: adding too many workers could lead to a situation where after this requester is done, we have nothing for them to do and they feel like the platform is slowing down or something.
2. Suggest the requester try a different option e.g. redesign the task, increase the price, or bridge to MTurk, this also has a downside in that the requester has already tried all of these options and so far none have made this task move fast for them. In other words, suggestion 1 is kind of like their last option.

At this point, I think I lean toward option 1 because I think it will help the requester and I don't feel that having a few more workers will be a significant challenge to us, even if we do run out of work for them for a short while.

The implications

These vary depending on how we choose to deal with this situation, but basically:

1. make a requester happy (they may be able to help us find others later)
2. get a few more workers (which is good for the short term, but could be a problem a little later if we don't have enough work for them)
3. these are pretty high value tasks as they take a lot of time, so they are quite valuable to workers, even though the tasks are not easy.

Person or People Who Added the Proposal to GitHub

The name in this section doesn't indicate that the person came up with this idea, unless it is explicitly and clearly mentioned.

@markwhiting added this to Github. Others are encouraged to contribute options and ideas about how we might deal with this and can be reflected here.

---

To officially join in, add yourself as an assignee to the proposal. To break consensus, comment using this template. To find out more about this process, read the how-to.

[dmorina]

Do you know how active are our 5 most recently added workers from the mentioned group for this particular task? Just trying to figure out roughly how many new workers we would need to add to achieve a 3x i.e what's our maximum before we start seeing any issues that you mention if any exist

Option 2 doesn't sound that fruitful especially when the requester has tried most of those so my vote would be to carefully execute option 1 by gradually adding new folks and explaining that the number of tasks on the platform can be low

[mbernst]

I wonder if we might recruit by letting workers know the general parameters of this task (e.g., how long it takes, how difficult it can be, the volume of work the requester is looking for). That way, those who volunteer might be likely to stick with it.

[markwhiting]

My thoughts on concrete next actions on this are:

1. let the requester know we are going to try to support them with more workers. @mbernst do you want to reply because it was mailed to you? If not, I can with the support address.
2. add a few workers at a time. I have a list, I can't add them remotely though (but will be back from the 1st so can then). Perhaps @dmorina, can I send you a list to start with before that?

To @dmorina's point: I don't have exact numbers now but the latest batch were our best set as far as I know.
To @mbernst's point: This is a nice idea, but in practice it seems that our pool is sufficiently excited to try that I don't think anyone is going to say no.

[qwertyone]

Here is a wild idea. What if there were a Daemo Token system to accompany the normal payment? I wonder if a task based merit system might create a useful value for the workers. The immediate impact of such a thing would be an increased cost of operating a Daemo system, as well as the normal development costs. In the long term, it would begin a second economy that might grow atop the normal system.

Of course this is a typical response; I think the value would be driven by how the workers could use the created system.

[mbernst]

@markwhiting I don’t feel that I need to respond. Using the collective support address is probably the better option.

[markwhiting]

@qwertyone token based system design is tricky but interesting. Lets discuss this separately, to see if there's an interesting road forward.

[markwhiting]

Just an update on this. 5 more are added and the requester was notified. I will monitor the progress over the next 2 days and add some more workers later in the week responding to that.

[shirishgoyal]

@markwhiting Do we have answer to below before executing it? Infact we should know the frequency distribution for the number of tasks completed by the workers.

Do you know how active are our 5 most recently added workers from the mentioned group for this particular task?