My *opinion* is that **they probably do not help, but that could change if you post a large enough bounty and are somehow able to "advertise" it (maybe in chat) to enough people that can answer** (or upvote your app) and who are not regulars on Stack Apps, but willing to join the site.

There are very few people who regularly visit Stack Apps.  Of those, I estimate that maybe 5, or so, will answer a general question with any regularity.  And frankly, would probably answer it anyway, if they thought it would help -- no bounty needed.

##Criteria to measure Bounty performance:
What positive effects would a bounty ideally have?  
Here are some possible benefits:

 1. Increase the **number of views** to a question.
 2. Increase the **number of answers** on a question.
 3. Increase the number of **"new" users** viewing/answering a question.
 4. Increase the odds of an **accepted answer**.
 5. **Awarding full bounty** to the answer the bounty provider most likes.
 6. **Awarding partial bounty** to a well-received answer if the provider does not make the award.
 7. Increase the **number of votes** on the question and its answers (often a function of higher viewcount in a short amount of time).

##What does the data say about Bounties on Stack Apps?

See [**this SEDE query**][1]. It shows:

 1. For the entire existence of Stack Apps, there have been **only 23 bounties offered**.
 2. One of those is on a deleted question, so further information on it is limited.
 3. Only 7 of the remaining 22 have an **accepted answer**.
 4. 5 of the questions have **no answer**.
 5. 11 of the questions **have only 1 answer**.
 6. **Only 10 had the bounty awarded** (so 12 had no "worthy" answer added by the bounty).
 7. One of those bounty awards was **not for the full amount** -- suggesting the award was automatic and the bounty provider was not happy.
 8. 3 of the bounties were awarded in 3 days or less. This is usually to **"reward" existing answers rather than to seek new ones**.
 9. The **average question score** was 12.1, but only about 3.9 for questions opened since 2013.
 10. The **average view count** was 770,  but only about 358 for questions opened since 2013.  
  

----------
Note: Those view and vote figures should be calculated relative to non-bounty questions and to how many days old a question is and what date the bounty was placed.  I'll leave all that as a future exercise.

  [1]: https://data.stackexchange.com/stackapps/query/886014/list-all-bounties-by-their-question-with-outcome-awarded-views-num-answers-e