For articles without an illustration, an algorithm (potential) suggests an image from Commons. This might be a simple algorithm that just looks at what images are used on that article in other languages. The newcomer decides if the image belongs, and where in the article to add it.
Good images make a big difference in an article, and newcomers are interested in adding images.
Adding the wrong image to an article could damage the article in a very visible way.
Add a reference
Some sentences or paragraphs clearly need citations. An algorithm (in development) would point out which sentences likely need suggestions, and the newcomer would seek sources to add as citations in a step-by-step workflow.
References are of clear importance to the core of the encyclopedia.
This task may not be exciting to newcomers. They may also struggle to find and use sources without guidance.
Using open-source spellcheck dictionaries and code, or using Wiktionary, identify likely misspelled words, and point them out to newcomers, who can use the visual editor or wikitext editor to fix them one at a time.
Clearly valuable and needed in any wiki, satisfying to newcomers. Helps them start editing the main text of articles, as opposed to peripherals parts of the article.
Scaling to any language may be difficult, depending on the availability of good spellchecking algorithms.
Add a section
An algorithm detects when an article could use additional sections, based on the kinds of section headers that similar articles have (e.g. all biographies of scientists tend to have "Publications" sections). The newcomer is walked through producing a well-referenced paragraph.
Real content additions that could help close knowledge gaps.
A much more challenging task than the others, requiring many wiki skills to be used together. May produce low-quality content.
Prioritizing "add a link"
The Growth team currently (May 2020) wants to prioritize the "add a link" workflow over the other ones listed in the table above. Although other workflows, such as "copyedit", seem to be more valuable, there are a set of reasons we would want to start first with "add a link":
In the near term, the most important thing we would want to do first is to prove the concept that "structured tasks" can work. Therefore, we would want to build the simplest one, so that we can deploy to users and gain learnings, without having to invest too much in the first version. If the first version goes well, then we would have the confidence to invest in types of tasks that are more difficult to build.
"Add a link" seems to be the simplest for us to build because there already exists an algorithm built by the WMF Research team that seems to do a good job of suggesting wikilinks (see the Algorithm section).
Adding a wikilink doesn't usually require the newcomer to type anything of their own, which we think will make it particularly simple for us to design and build -- and for the newcomer to accomplish.
Adding a wikilink seems to be a low-risk edit. In other words, the content of an article can't be as compromised through adding links incorrectly as it could through adding references or images incorrectly.
Notes on "copyedit"
In conversations with community members on this project's discussion page, many people brought up the question of how to make a structured task around copyediting. Correcting spelling, grammar, punctuation, and tone seemed to everyone to be a clearly useful task that should be prioritized. The Growth team initially shied away from this workflow because of scaling concerns: even if we were able to find or develop an algorithm that could reliably find copyedits in one language, would we be able to do that in dozens of other languages?
We began to learn more about this by talking with User:Beland, who developed the "moss" script for English Wikipedia's Typo Team. We wanted to understand how the process works, and what it might look like to do something similar in other languages. In short, it sounds like the most promising avenue is through existing open-source spellcheckers and dictionaries. Two examples are the aspell and hunspell libraries. Below are our notes from learning about "moss" with User:Beland.
Prospects for doing something similar in other languages
A process like this should theoretically work in other languages, given that other languages also have Wiktionaries and open-source spellcheckers.
But it would not be possible to deploy in a new language without native speakers validating it. There would likely need to be customization for many languages.