Using Twitter for Storytelling

robert | March 10, 2011  ‐  5 Comments

There’s an excellent post on the problems and experiences of Twitter storytelling at Sliverstring Media but while I’m waiting for my comment to be moderated I thought I’d re-blog it here:

The key to success with storytelling in any media is to work with the strengths of the platform. Twitter is a real-time, social, conversational stream that is best used to invite and build participation. Thinking of Twitter as thousands of 140-character “book pages” is the wrong mindset. It’s like thinking that a short story is just a long story with fewer pages or a short film is a 15 min feature film.

The key to Twitter storytelling is:

(a) use it to invite participation. Create scenarios and “exercises” that open the door to followers to contribute. Make it conversational. Allow followers to become advocates by facilitating the spread of the participation, not only the spread of the tweet. That is, it’s not simply a RT of the story tweet but an invitation from one follower to a non-follower to get involved – perhaps using some game mechanics with the storytelling to provoke and reward that.

(b) recognize that Twitter is both a Discovery and an Exploration platform. That is, current & recent story tweets and the participatory tweets are Discovery content – they’re luring audience into the world. At the same time the historical Tweets offer backstory and context – Exploration content – for those in the  audience that want to dig deeper. Hence you’re right that audience should be able to dip in at any time in the life of the story and become immediately engaged without having to read the premise/synopsis etc. The way to achieve this is to finely craft each Tweet so that it works like a Zen koan – it’s a 140 character meditation on the story that is revealing, intriguing and surprising. This is particularly important if the tweet is from the voice of a narrator rather than a character. I have always measured the strength of a short story by whether it leaves me thinking about the premise of the story for longer that it took to read. The same should be true for every Tweet. Remember that twitter is a real-time news stream which means you’re only as good as your last tweet

(c) use it to build & populate the world. As I hinted above, a story might have several Twitter streams from the perspective of different characters or entities. This means that while a “narrator” stream might tell *the* story, other streams might shed new light and different perspectives on the narrator’s voice. As with any transmedia experiences, these new streams should all add value to the core narrative yet at the same time be optionally consumed. One example I’ve been exploring with a storyteller is to have a twitter stream for a fictional Government bureau in much the same way as George Orwell has in 1984 – the  stream sends continual optimistic official news  “production up by 120%”, “inflation static at 1%”, “crop yield the best since records began” – which is directly contradictory to the experience of the narrator! Such a stream builds out the world with a new richness but is timed to impact the through-narrative should someone choose to read both. I appreciate that this may contradict the “Zen koan rule” but then it’s not being used for Discovery, it’s Exploration so I’ll allow myself some latitude ;)

In terms of commercializing the Twitter platform, it’s value is in the social spread of the story and the building of audiences. Revenue should be taken from other platforms.

Calls to participate “case (a)” are much easier to provide examples for than the koan “case (b)” although you’ve listed some good places to research.

Jay Bushman’s Twitter stories are always provoking and inspiring followers to create their own stories. He brings the fictional setup, let’s say the context or the world, but then it’s up to everyone else to bring their imaginations and participation.

For #sxsw we’re running a rather trivial story of the Three Pigs by way of illustrating the mechanism of participation and interactive narrative. Firstly we stage the story as a competitive game between the pigs and the wolf – the battle outcome determining the course of the story – and secondly we’re using tweets from the pigs and wolf to provoke reaction and participation from friends and followers. Using our Conducttr platform we can facilitate some of the invitation to participate using our “3rd party reply” feature which takes a follower’s friend’s Twitter ID and sends it a message from the fictional character. What we’re doing is not meant to be a gold-standard example of this thinking/storytelling in action but a simple eye-opener.

  • Pingback: Twitter como herramienta para contar historias | Noticias TransMedia

  • http://twitter.com/transistoria transistoria

    I recommend you http://365musicaltweets.com/, another way to tell stories with Twitter and music ;)


    The project “365 musical tweets” is a musical action. It consists in publishing, via my twitter accounts (@MikelChamizo and @365MusicTweets), a daily musical tweet from the 1st of January to the 31 of December 2011.

    Each tweet has a maximum extension of 140 notes or chords and it responds, like it occurs with text tweets, to and idea, occurrence, reflection, experience, motivation or news that has caught my attention that day. It’s a diary, in essence.

  • Guest

    Can you explain how you used the Conductrr platform with The Three Pigs? I see how you did a very attractive blog, and how you did a series of Tweets. But, what I don’t see are Tweets that ask for a response of an answer… ex. (a) or (b) that would trigger a response which is what I understand Conductrr can do. How do you capture the follower’s friend’s Twitter ID? Does your follower just give you their Twitter ID? Or do they need to be following you? In that case, wouldn’t the follower get your Tweet anyway? I would love to see you take this case and do a series of video tutorials on Conductrr using this story as an example.

  • http://www.tstoryteller.com robpratten

    For The Three Pigs, Conducttr did the following:
    (a) manage the blog posts
    (b) manage the interactivity.

    A. Blog Posts.
    The story as I implemented it had a branching narrative each time the wolf approach a house. So:
    i. at the straw house the 1st little pig could be eaten or he could run away.
    ii. at the stick house the 2nd little pig could be eaten or run away BUT the 1st little might be there too if he survived the first attack
    iii. at the brick house I turned the tables a little and put the threat on the wolf – he could walk away or get eaten.

    I use an unlisted YouTube videos to represent the fate of the pigs & wolf
    Video1 = pig 1.
    Video2 = pig 2
    Video3 = wolf

    If the video was unpublished (e.g. private) then the pig was alive, if it was published then the pig (or wolf) was dead. In this case the videos were never shown to anyone because they were just used as flags or markers but they could have had animations or live action of the pigs dying or escaping.

    So, the story was broken into three acts lasting 20 mins each.
    Act 1 – the straw house
    Act 2 – the stick house
    Act 3 – the brick house.

    Every five minutes a new blog post was posted. Conducttr had all the content and the Date & Time trigger was used to advance the story.
    So, 5min = blog post 1
    10 mins = blog post 2
    15 mins = blog post 3
    so far so good… but now at 20 mins we need to see if the first little pig has lived or died!

    For this a Boolean Trigger is used with the Date & Time that basically says IF the time is now 20mins AND Video1 is public THEN tell the story about how the first little pig got eaten.

    The decision to make the Videos public or private was taken care of in the game so..

    B. Interactivity
    In this example Twitter was used to build participation and discovery around the story. It was not necessary to follow any of the characters and building a following was not a success criteria – the goal was getting people at SXSW to attend our presentation.

    With this in mind we used:
    i. the 3rd Party tweet that allows a friend to poke the story character into tweeting their friend. This is shown in the SWSX presentation.
    ii. simple Reply tweets to “huff”, “puff”, “blow”, “bacon”,”help”,”run” etc. The tweets were intended to be humorous and fun.
    In Conducttr we used several Twitter Triggers on all the characters and scripted replies to any incoming tweets.

    The characters also had some schedule (Date & TIme) tweets and Facebook status updates that were delivered during SXSW.

    I hope this helps?


    ps. this is a cut and paste from the Synopsis of the The Pigs Project. Because I was using the project to test Conducttr I’ll admit it’s over complicated :D But it did work!

    The Three Little Pigs
    Written for demonstration of interactive storytelling for SXSW 2011

    Three pigs are:
    3rd_Little_Pig and

    A. Create all the logic – SECTION 1
    B. Create the content – SECTION 2

    Runs between 12:30 and 13:30 on March 15th

    ** 5 Minute Triggers **

    D1 = 12:30
    D2 = 12:35
    D3 = 12:40
    D4 = 12:45
    D5 = 12:50 —> DECISION POINT 1 on PIG 1 LIVE or DIE —> P1 = Pig1 DIES
    D6 = 12:55
    D7 = 13:00
    D8 = 13:05
    D9 = 13:10 —-> DECISION POINT 2 on PIG 2 LIVE or DIE —> P2 = Pig2 DIES

    ** Decision Points, set using Twitter **
    Tweet from puppet master decides if the pig lives or dies. Since a Tweet is a repeat trigger, use it to publish a dummy video that will then stay true.

    T1 = Pig1 = Pig 1 is DEAD –> publishes Video One —> P4 used to test video 1
    T2 = Pig2 = Pig 2 is DEAD –> publishes Video Two —> P5 used to test video 2
    T3 = wolf = Wolf lives on –> publishes Video Three —> P6 used to test video 3

    ** Booleans ***
    “ACT 1″ – house of straw
    D1 to D4 –> no booleans required

    “ACT 2″ – house of sticks
    story if PIG 1 is DEAD (e.g. P1)
    B1 = D5 AND P1
    B2 = D6 AND P1
    B3 = D7 AND P1
    B4 = D8 AND P1

    story if PIG 1 is ALIVE (e.g. NOT P1)
    B5 = D5 BUT NOT P1
    B6 = D6 BUT NOT P1
    B7 = D7 BUT NOT P1
    B8 = D8 BUT NOT P1

    “ACT 2″ – house of bricks
    This comes after P1 so combine the two
    (a) Pig 1 and Pig 2 are both dead – both killed in the house of sticks
    (b) Pig 1 is dead and Pig 2 is alive
    MISTAKE! FORGOT: (c) Both pigs LIVE
    B30 = D9 BUT NOT P1
    B31 = B30 BUT NOT P2 = Both alive and run to sticks

    story if BOTH PIGs are DEAD (e.g P2 and P1)
    B9 = P1 AND P2 (both pigs are dead)
    B10 = D9 AND B9 (decision point and both pigs are dead)
    B11 = D10 AND B9
    B12 = D11 AND B9
    B13 = D12 AND B9

    story if ONLY PIG 1 is DEAD but PIG 2 LIVES (e.g P1 BUT NOT P2)
    B14 = P1 BUT NOT P2
    B15 = D9 AND B14 (decision point and pig 1 dead but pig 2 lives)
    B16 = D10 AND B14
    B17 = D11 AND B14
    B18 = D12 AND B14

    P3 = WOLF lives – e.g. he doesn’t go down the chimney

    (a) all three pigs are alive and wolf gets eaten
    (b) pig 1 died but pigs 2 and 3 are alive and wolf gets eaten
    (c) all three pigs alive and wolf walks away
    (d) pig 1 died but pigs 2 and 3 are alive and wolf walks away
    (e) pigs 1 and 2 are dead but pig 3 lives and wolf gets eaten
    (f) pigs 1 and 2 are dead but pig 3 lives and wolf walks away

    —->> pig 1 alive = D13 BUT NOT P1; pig 2 alive = D13 BUT NOT P2; wolf walks away = D13 AND P3

    B19 = D13 BUT NOT P1 [PIG 1 ALIVE]
    B20 = D13 BUT NOT P2 [PIG 2 ALIVE]
    B21 = D13 AND P3 [WOLF ALIVE]

    (a) ALL pigs alive and Wolf gets eaten == B23
    B22 = B19 AND B20 (all pigs live)
    B23 = B22 BUT NOT P3

    (b) pig 1 died but pigs 2 and 3 are alive and wolf gets eaten == B25
    B24 = P1 AND B20 (pig 1 dead, other pigs alive)
    B25 = B24 BUT NOT P3

    (c) all three pigs alive and wolf walks away == B26
    B26 = B22 AND P3

    (d) pig 1 died but pigs 2 and 3 are alive and wolf walks away == B27
    B27 = B24 AND P3

    (e) pigs 1 and 2 are dead but pig 3 lives and wolf gets eaten == B28
    B28 = B9 BUT NOT P3

    (f) pigs 1 and 2 are dead but pig 3 lives and wolf walks away == B29
    B29 = B9 AND P3

    ** INTERACTIVE Booleans ***

    Goal: send different tweets depending on status of Pig

    T08 = “run” matchword
    D5 = False = respond to “run”

    Three states:
    a. pig is alive in own home
    b. pig is alive in brother’s home
    c. pig is dead

    a. B33 = T08 BUT NOT D5 (run before the decision point)
    c. B35 = T08 AND P1 (pig is dead)
    b. B34 = T08 AND D5 (run after decision point)
    B36 = B34 BUT NOT P1 (after decision point and pig not dead)

    a. B37 = T09 BUT NOT D9 (run before the decision point)
    c. B38 = T09 AND P2 (pig is dead)
    b. B39 = T09 AND D9 (run after decision point)
    B40 = B39 BUT NOT P2 (after decision point and pig not dead)

  • Pingback: continue