NN/g Nielsen Norman Group Login
Wot Leadersin Research-Based User Experience Ea
Home Articles Taming Events Consulting Reports &Books About Nig
‘Topics Information Foraging: A Theory of How People Navigate on
sie
eign rocess the Web
Ecommerce ‘Summary: To decide whether to visit page, people take into account how much celevant information they
tore arelilely to ind on that page relative to the effort involved in extracting that ino,
Navigation
Peyhology nd Ut also Bs Rover 10,2019 Sete
ae “ope: Pico ond UX, veto vine
study Guides
Users ‘it people sro ona webpage? How do they decid a cek ona nk? Vien do ey leave a webpage? When do ey
‘Web Usabitty prefer to search and when do they browse’? How do they decdeto search fr ifomation ina moble app or onthe
etingtorthe web web?
Se altorce ‘These and many other questions about websser behaviors can be answered by the informationoraging theory.
Popular articles Inthis arice we presen abroad ovenvew of the theary and review some of ts imlcatons fr web desion
a. Usabity Heuristis for User
Inerace Design
Empathy Mapping: The Fist Step n
Design Thinking lnformation foraging was deveiones a PARG (former KEROX PARC) by Peter Pl and Stuart Gardin he late 1980
‘Went Use Which User perience and was inspired by animal behavior theories about how amas forage for fod (hence the name). Thus, not
Research Methods surprisingly, animal foraging ane information foraging thectles share comman tetminology, a sted inte table below.
Analogy with Animal Foraging Behavior
Service Blueprint: Defrition
paaaaneinehey ANIMAL FORAGING INFORMATION FORAGING
The Four Dimension of Tone of Voice
Between Subjects. Within Subjects
Study Design|
Uc Researeh Cheat het oe
sbi 2: ndation Lab,
Usabity Testing 101 1:9)
recent Articles SS
ey tis mata gen path wi
Lest Sage: erat ae
Cones Method: stay uide
How to Handle Category Outliers in cy Teta aod nee atan
voor anal may coir order
rroimecuntanesanuoe BB Saying
PorcePincle
‘2 Moysto courage Employee Sharing
and Engagementon an insanet What Is Information Foraging?
necent Videoe Iromaton raging the fundamental theory how pepe navigate on the web sty an ration nee
ssn sys at von uses hav a eta intomaton goal, ey assess he oration tat hay can extrac om
_3Woysto Level Your vl Design says tha, oa, ey me
any ‘any cand source of ivomation rave oe osnvledn exacting Ht noxratn and hack on or
vepvowwhoeigatntenge2nn SEI CAMBSeSOUcEs oat hey maximize the rato:
Product UX Parr. Rat of gun = Information valu / Gost associate with obtaining that information
Fetemetcusity 094 gp ya peop hare a question hy wl desde we webpage got ase on (owe it ale
Functionality age wl provide an answer to ther question, and (2) how long t's going to take to get the anster they goto thatbeter Xx Deliverables
Authors
Bruce Tognazini
on Norman
See all authors
New atu event: Intranet
Employee Experience sympesirm
Live Novernber 10, ith winners
fof thentanet Design Annus
contest
page,
(Anima behavior sclence shows tht imi typeof optimization hale tue fr animal foraging — hence, the opti
{foraging neory hat Served as source of inspiration fer Pol and Card. Basically, an animal needs to eat more calories
‘than expends, or wil starve and ultimately ce witout ospring Across many generations, animals have evoWved
highly optimize foodoraging strategies)
Intayman terms information foraging explains why neon dont scrll minglesaly or elk on every single ink on the
[pages because they attempt to maximize the rate of gain and get a much relevant information in as lite time ae
[possible. Scrolng o clicking aot more would probably gain Ihe user mor information, Duin the user's estimation,
{he rte-o-gain rato would decrease, because the numerator (te iformaton value) wou incease oo tle compared
{othe increase nthe denominator (the interaion cost associated to geting the infomation).
‘You may wonder how people can be so rational as to always take the action that maximizes their gain. Aral, we have
‘counties examples of people behaving rational, against thelr oun interest. fac, umn behavier canbe well
‘descrives by what Nobel winner Herbert Simon called bounded rationally, Wheteas the chlces that people
‘make ater o maximize benef and minerize cost, humans have a har tne preitly estimating benef and cost
‘anc thus use salsfking and cher imperfect heussts to pick the most promising choices. The overal effects stl that
‘of siming for he highest rate of gain, even ifthe achieved number may not be the theoreti! optimum in every eae.
‘Specialy, people hae no way of knowing in advance (1)how much formation a pach contains; (2) how much time
[wil take tem o extract nat information. Fora given task, they do know however how much real tie they spent so
{ar and ow much relevant conten they were able to get. n order fo choose what source to look a net, they estate
how the rate of gain wil change they choose to explore a speci information source, Ofcourse, te judgment wont
be perfect — iil simply be based on whatever external cues the information source emis
Information Foraging:
‘Choosing a New Info Patch
me
When peopl att to acquit internation for an information ned, the oe look at mult sources of information or
lformatonpches) Aton pint ntine they kaw the real infamation value they have goten oar femal the patches
they alrecdy vised and os the ea ot (or tims) pent to gather tht information. Tedecdewhsthero new otchis
orth exploring, the estimate how te rato information goin wll change they choose that patch, The estimate based
the on cue ha tey rece fam the patch about te nfo valueo that patch information see andthe pares eort
needed to extrac thot, ls possible tha these estates oe infact wrong and ha the real infermaton vue adr
‘asseited with patch re dileent than the user peeved infmatian value (a ifomation scent) an fo.
Information Scent
How do people estimate how much information they wil get om a page Before visting I? That's where the concept ot
information scant comes into play.
“Animals deci which pach to frage on based on (among cher things) the scent that hey catch fom the envionment:
Ite scent signals the food thatthe animal's interested int may decie to pursue h.
‘Similaty 253 user searches for infrmation onthe wob, she judges the webpages she encounters based on how
‘well suited they ae for her goal, Each source oinformation tus emis “scent —a signal tat tel the frager how
ty ts that contains what she needs,Note tha. the nthe animaboraging world where diferent animals eat iferent foods, the scents high depending on
‘the typ of information that he users interested in. Thus, a camivere may be arocte by the scent of blood, whereas
“an herbivore may be completly insonsiive tot and maybe peter the scent of pe Mult or ren gras. Similar, the
same source of information can have a great scent for ene user because tlds exactly what he’ inlerested in, Bu
ero Scent ora ferent user wth 2cferent information need,
‘What makes the scent of webpage? When a person lands on that very page, the scents given by the tte, Images,
andthe Information that fe easy vibe above the fol. the users searching for dish towels and lane on a site wih
Pictures of stawbertes, bear, and candy, she may assume that this page is unkely to contain what she needs simply
because the scent points no a erent dvection,
When someone leekng fo sh towels ands na page wthpltwes of candy bee, and trowberies, she may justia tink
that she nae Find what she needs there, simply becouse the information sent cues pont ina diferent dvecin.
‘wana person looks ata tnk oa paga, the scent I given by al the words and images associated wih that ink. Thus,
the same person tookng fresh owels may be stongly aac by a Iink called Ktzhen tren next to an apron, 3
chen glove, and several ash towels
CLEARANCE
Just For You
The ink Kitchen linens ond the asecatedimage hoe strong infomation scent forthe tsk of leokng For csh towels
Costs of Addressing an Information Need
Remember that he ae of ain equation above neues to diferent variables the information valve and the cost of‘obtaining the information. There ae two types of costs associated wth obtaining information (1) he actuate and
‘fort invclved in extracting the information from the various Information sources, and (2) the oppertunty cost —
resulting tom foregoing he Benet of exploring oer documents in favor ofthe chesen ones.
(ina for-pay system, there wilbe a thr cost in trms ofthe monetary price ofeach document, but since most webstes
ate tee to access, we wont lscuss payments further inthis are. The important pins tha users! time and masse
‘are real ost, even they don't come wth dolar signs attaches)
‘Opportunity Costs
‘Wmenever users decide to inspect a webpage they potently lose the opportunity of ooking at something ese, That's
‘iy makes sense that once users have landed ona page, they quick extract the gist without spending tme delving
Int the tals. Scroling down and reading every single wer on a page Is counterpeductve — the uty derived fram
‘such an action kel o be too small, whereas moving on and surface.scanning a dierent page may prove way more
proftabe
‘Time Costs: Between-Patch and Within-Patch
“The te reflects the eo involved in gathering he Information thatthe user needs.
Lokng up information on the Web oF on 8 webpage usualy Involves two lypes of user actions:
4. Between-patch acts: Gathering Information scurces (Le, patches)
2. Wiin-atch actives: Inspecting each pach to extract information for it
For example, you went to research teatments othe common ced, you may stat by tunning@ serch ona search
‘engine. The searches page wl er a set of koks tat cous be potential sources of information, A tis stage,
_youie working between patches, gathering information sources that you wl explore late,
‘once you have your search results, you wil neat clek on the most promising links and read through the content 0
‘extract relevant information. this stage you are working within the Indu panes, nepectng them an extracting
‘he information thats relevant to you,
Both between-patch and within-patch acvies conbute to the overall ime to satya information need anc both eon
Impact the user experence n many ways, Users can ceate adapatons that alow them to minimize te ume spent
between patches or tin patches. These adaptations are calles enrichments,
Enrichments.
oh the bebveen-patch ime an the wiinpatch time can be minimizes by enichments.
‘An enrichment roferso a user interaction, behavior strategy that aims 1o maximize the
Ut of te information feragng. can happen ether between patches of win patches
“Think of ervichments as an extra too! hat users may use in order to forage more efclenty, The tool may be something
‘that they sready haven ther pocket uch ae leaned behave) oF coud be Bul onthe spot and tala othe
‘specif pate (in whieh case the user must spend me eating)
[Behavior enrichments re the tole that users already have aequed and that hep them extract norman ein.
“These behavior are adaptations tha evolved overtime and that proves tobe success in many stvaions nthe past.
For example, over many interactions with tie web, users have devised behaviors that minimize the me spent gabeting
Patches or estimating information value of patch. Thus, inorder to avoid contextswiching costs, milennials use page
pathing te clearly separate he formatlr-entraton frm the iformatlon-gahering componens of the foraging. Orn
‘1der 1 save time, users use Fpl scanning an web.search results tobe ale to qucly assess scant wnout
reacng the ene test associated win a search result
‘Attn patch level, people have devised other caning pattems (eg, layer cake) to quicky late the relevant content,
“They also tend to ignore banners or he right rall— another adaptive Behavern serie of masimizing the Information
‘gain relative tote me spent onthe page.
Interaction enrichments require extra effort rom the user — these af the to's tha the user needs to bull onthe
‘pat, n erder to maximize he efcincy ofthe information foraging ask.
For example, he user can spend te ta think of spect Keywords that best describe her query, hoping to increase the‘win patches, people can also use strategies such asthe use of wthin-page search to quickly locate content thats
relevant to hers
Good User Experience Can Make Enrichments Unnecessary
Enrichment ae shy forthe use foro reasons: nt nly isthe case that some enrichments equie users ta pay an
‘otra ineracton costo ake upon iemselves larger cognive oad, but there is a chance that they wont be igh for
{he task For example, a user using an Fpattern to Sean a webpage may miss important concepts that dont appear at
{the beginning ofa paragraph. Cx, a person searching for her fst lawn mower may not have enough domain knowedge
‘ocome up witha set ofrelevant evans in serch fo.
"Now, imagine that the between-patch and the wihinpaich environment waa one sep ahead of users and adapta
‘homselves to user needs So tal the user dan need to create any special enichments: these envionments were
‘aleady lat out to maize the efficiency of information foraging. Thats what a good user experience Is
‘Agood user experience inves web pages Dat are designed so thatthe user can get the maximum relevant
Infomation nthe minimum amount of ime
‘Some optimization fs aiteady in pace. Indeed, today's search engines use sophisticated ways fo rank the search resus
‘50 hal the most relevant ones appear fist. They use common searches and personalize thelr resus so that even the
user doesn't take the time to refine her query there Is a high chance o get what she wants because the search engine
knows" her or because that's what mest people mean when they run a similar query. Aulosvagestion for searches also
reduce the time of thinking about hese enichments.
‘Ave page level, the way n which information is ad out and presenied on a wel-esigned page is suggestive of
‘whats most ikely to be relevant ona page. For example, designers use scanning trendy formating suchas bulleted
lists, bolded keywords, descriptive headlines io help users id the information that is celevant to them.
‘Of course, the tet prblem Is: each usr i looking for ferent type of food. She has a certain very spectc
Invomaton goal in mind that may be deren than that of your next user. How do you optimize your page layout and the
Information scent hat yout page ents so that it works for all users and or al goals?
“The ansne is: you dont. You opinize forthe top task tat your page or ste Is supposed to adress. That wi take care
‘ofthe many users who wl be dong that as, Iwi also take cre ofthe many users tempting to solve aoferent
1tas, for wham your page is ielevant — they write tempted tock on your page
Conclusion
_ymen people lk for information on the web, they attempt to maxsnize te at of information gain over te: they want
to getas much information as passe Inthe minimum amount of ime, To resue time, hey estimate he information
value ofa page based on is Iformaton scent and use evichments such as larned behaviors o Interactions to
Improve the chance o quik geting what they needin thei information foraging.
Reference
P. Proll, S. Care, 1899. Information foraging. Peyehological Review 106, pp. 649-675.
Downloads
Information oraging tustratons (A4) POF)
Infermation foraging Hustrations (etter) (POF)
Learn More
‘Subscribe to the weekly newseterto get notified about future aces.Asticles UX conference Training Courses
Cancel ve Close: Dealgnto Distinguish te Diference Pers ussive and Emational Design
Information Foraging: Why Google Makes People Leave The Human Mind and Usability
‘Your Ste Faster
Hour Talks
“The Peak-End Rule: How impressions Become
Memeries ‘Search and Findabity
“Te Anchoring Principle Helping Intranet Users Fina What They Need
Distracted Driving: UXs Responsibly to Do No Harm
Research Reports
Universy Websites
Vol 07: Navigation and Page Layout
Inanet Information Architecture Design Mathes an
Case Studies
Videos
‘Short-Term Memory mations Working Memery ana External ‘Sorta Poot UX:
Impact User Interface Design Memory 2 minute veo
minute video ‘minute video
Popular Topics About Follow us
pgteetenn ux eaies Tablet Design ontitormation sebsebetw ou
>oplcation son Persuasive Desion Aeoets
Design Process search Methods Why WO WD Fotton nngon Twiter