Each panel below is a working miniature of a site that could run at this address: navigation, hero, product screen and proof numbers, followed by a plain note on how that business would position the name. All twenty stay on screen at all times; a lane simply marks its own. The domain sits in the browser bar of every one of them, whole and unbroken, because the name is the thing being sold.
01
The Hour Is Measured
The federal government publishes the curve
MathSolverNow
CurveHoursMethod
See the curve
Four point six
at eleven. Point four
at eleven.
The Bureau of Labor Statistics publishes one table showing the share of Americans doing each activity by hour of the day. Educational activity peaks late morning and falls all evening. It does not reach zero.
ATUS TABLE A-3 · EDUCATIONAL ACTIVITIES · 2025PERCENT OF POPULATION
PEAK 11 A.M.AFTER 6 P.M.DATA REFER TO PEOPLE 15 YEARS AND OVER
2025 ANNUAL AVERAGES, PUBLISHED JUNE 2026A GOVERNMENT TABLE
4.6%at 11 a.m.
1.6%at 6 p.m.
0.4%at 11 p.m.
How they position itA category that sells availability should open on the only federal table that publishes an hour-by-hour curve. Nobody in this market cites it, and it is free.
02
When the Day Ends
Start time plus day length, and the arithmetic is ours
MathSolverNow
StartLengthAfter
Do the arithmetic
School ends
in the afternoon.
The Department of Education publishes an average start time and an average school day length, from two different surveys in two different years. It publishes no average end time, so the end is arithmetic and this page labels it that way.
PUBLIC HIGH SCHOOL, TWO NCES SOURCESDERIVED
AVERAGE START8:00 a.m.NTPS 2017-18
AVERAGE START8:07 a.m.NTPS 2020-21
AVERAGE DAY6.7 hoursSASS 2011-12
MIDNIGHTIN SCHOOLAFTER
OUR ARITHMETIC, NOT A PUBLISHED FIGUREENDS ABOUT 2:45 TO 3:00 P.M.
NCES PUBLISHES NO AVERAGE END TIME FOR PUBLIC HIGH SCHOOLSSO WE SHOW THE INPUTS
8:00 a.m.average start
6.7 hoursaverage day
No figurepublished end time
How they position itEvery product in this market is used after a bell rings. Showing the inputs and calling the answer arithmetic is what separates a reference from a brochure.
03
Two Populations, Same Clock
One is at work, one is at homework
MathSolverNow
CompareHoursSource
Compare the rows
At ten at night,
3.6 against 0.4.
The same federal table carries both series. Read the working row and the educational row at the same hours and the evening is the whole story: the staffed world empties out, and the schoolwork does not stop with it.
ATUS TABLE A-3 · 2025 ANNUAL AVERAGESTWO ROWS
WORKING AND WORK-RELATEDEDUCATIONAL ACTIVITIES
10 a.m.29.6%4.5%
3 p.m.21.6%2.7%
6 p.m.11.2%1.6%
8 p.m.5.9%1.1%
10 p.m.3.6%1.0%
THE SHARE AT WORK FALLS BY EIGHT TIMES ACROSS THOSE HOURSTHE OTHER ROW DOES NOT
29.6%working at 10 a.m.
3.6%working at 10 p.m.
Same tablesame hours
How they position itThe contrast is already in the government's own table and takes one screen to show. That is the kind of fact a reference product is built on.
04
Weekdays and Weekends
Different days, different amounts, same evening
MathSolverNow
DaysMinutesVintage
Split the week
Forty-nine
against thirty.
The most specific federal figure on high school homework is a 2008 visual essay drawn from time-use data. It is old and this page dates it, but it is the only published breakout by age that exists.
HIGH SCHOOL STUDENTS AGED 15 TO 19MONTHLY LABOR REVIEW, NOV 2008
AVERAGE WEEKDAY49%DID HOMEWORK1.7 hours on the days they did
AVERAGE WEEKEND DAY30%DID HOMEWORK2.4 hours on the days they did
Data period stated on the chart: September through May, 2003 to 2007. Weekday holidays excluded.
NO NEWER PUBLISHED BREAKOUT BY AGE EXISTSSO WE PRINT THE DATE
49%did homework, weekday
30%weekend day
2003-07the data period
How they position itDating a weak number instead of hiding it is the fastest way to be believed about the strong ones on the same page.
05
Three Years, One Shape
The decline repeats; the bumps do not
MathSolverNow
202320242025
Stack the years
The fall repeats.
The spike does not.
Run the evening hours of the same table for three consecutive years. The decline after mid-afternoon is in all three. The small evening bumps move around and are worth tenths of a percent, so this page does not build on them.
EDUCATIONAL ACTIVITIES · EVENING HOURSPERCENT
5p6p7p8p9p10p11p
20231.81.61.31.30.90.80.6
20241.81.71.51.81.30.70.3
20251.91.61.31.11.51.00.4
A BUMP THAT MOVED YEAR TO YEARHOUSEHOLD SURVEY, TENTHS OF A POINT
WE DO NOT CLAIM A NINE O’CLOCK SPIKEWE CLAIM THE DECLINE
3 yearssame shape
Tenthsthe size of a bump
Declinethe durable claim
How they position itRefusing to draw a spike out of survey noise is a product decision as much as an editorial one, and a buyer building a data page will recognize it.
06
Five to Ten, Sunday to Thursday
The best-known free math homework hotline in the country
MathSolverNow
HoursGradesCost
Check the hours
Free, staffed,
and closed by ten.
A hotline run by an engineering institute answers math and science questions for grades six through twelve at no charge. Its published schedule is five hours a night, five nights a week, during the school year.
PUBLISHED SCHEDULE · FREE MATH AND SCIENCE HOTLINESCHOOL YEAR
SUNMONTUEWEDTHUFRISAT
3 p.m.
5 p.m.
7 p.m.
9 p.m.
11 p.m.
STAFFEDNOT STAFFEDTWENTY-FIVE OF ONE HUNDRED SIXTY-EIGHT HOURS
THIS IS A REAL SERVICE DOING REAL WORK, FOR FREEIT ALSO HAS A CLOSING TIME
5-10 p.m.Eastern
Sun-Thufive nights
6-12grade range
How they position itThe strongest fact in this category is somebody else's published schedule. It is verifiable, it is not a criticism of anyone, and it describes the gap exactly.
07
And Closed for the Summer
The calendar is published too
MathSolverNow
CalendarBreaksSummer
Open the calendar
Closed the week
of Thanksgiving.
The same hotline publishes its closures for the year: a start date in August, the Thanksgiving week, a winter break of more than two weeks, and a summer that begins in the middle of May.
PUBLISHED CLOSURES, AS STATED BY THE PROGRAMONE SCHOOL YEAR
JULOCTJANAPRJUN
OPENCLOSED
THANKSGIVINGNOV 19 TO NOV 25
WINTERDEC 22 TO JAN 6
SUMMERFROM MAY 17
SEGMENTS ARE PROPORTIONAL TO THE PUBLISHED DATESNOT TO DEMAND
Aug 6opens
May 17closes for summer
2published breaks
How they position itSeasonality is the part of this market that nobody puts in a deck. A page that shows it is talking to an operator, not to a browser.
08
Three to Six, at the Branch
Public libraries, and what they can and cannot cover
MathSolverNow
BranchOnlineGrades
Find the window
In person until six.
Online until eleven.
Two large city library systems publish their homework help hours. The in-person program runs three hours on four afternoons for elementary and middle grades. The online tutoring runs longer, and that difference is the whole shape of this market.
TWO CITY LIBRARY SYSTEMS, PUBLISHED HOURSFREE
IN PERSON, TEN BRANCHES3:00 - 6:00 p.m.MONDAY TO THURSDAY · GRADES K-8
ONLINE TUTORING2:00 - 11:00 p.m.EVERY DAY · THIRD-PARTY SERVICE
12a6a12p6p12a
THE HUMAN WINDOW IS THE SHORTER ONEIN BOTH SYSTEMS
3-6 p.m.in person, Mon-Thu
10branch locations
2-11 p.m.online, daily
How they position itA product that wants library and district distribution has to know these hours cold. The name is the easiest thing in the stack to remember at the reference desk.
09
Three Hours After the Bell
The federal evaluation said so in its own words
MathSolverNow
CentersHoursDays
Read the finding
Generally open
for three hours.
The federal evaluation of the national afterschool program described the centers plainly: generally open for three hours after school, four or five days a week. That is the shape of the supported window, straight from the evaluation of record.
NATIONAL EVALUATION, 21ST CENTURY COMMUNITY LEARNING CENTERSQUOTED
Centers “generally were open for three hours after school four or five days a week.”
SCHOOLCENTERUNCOVERED
PUBLISHEDOCTOBER 2004
WHY WE STILL QUOTE ITIT IS THE FEDERAL EVALUATION OF RECORD
WE DATE IT BECAUSE IT IS OLDNOT BECAUSE IT IS WRONG
3 hoursafter school
4-5days a week
2004the evaluation
How they position itInstitutional buyers know this program by name. Meeting them with its own sentence, dated, is how a reference brand earns a first meeting.
10
Who Actually Answers
A university, a nonprofit, and two city libraries
MathSolverNow
ProgramsHoursGap
See the list
Four programs.
Four closing times.
These are the real, staffed, free programs with published hours that a search turns up. Every one of them is admirable and every one of them closes. No state agency runs one with published hours that we could find.
FREE HOMEWORK HELP WITH PUBLISHED HOURSWHAT A SEARCH RETURNS
PROGRAMRUN BYHOURSDAYS
Math and science hotlinePrivate university5-10 p.m.Sun-Thu
State homework hotlineNonprofit4-8 p.m.Mon-Thu
Library, in personCity library3-6 p.m.Mon-Thu
Library, onlineCity library2-11 p.m.Daily
THE HOURS NOBODY PUBLISHES A PROGRAM FORAFTER TEN, AND ALL WEEKEND
A 211 LINE IS NOT A HOMEWORK RESOURCE; ITS OWN SITE LISTS HOUSING, FOOD AND HEALTHWE CHECKED
4programs with published hours
0run by a state agency
10 p.m.the latest close
How they position itMapping the free layer honestly is how a paid product finds the hours worth charging for. It also tells a buyer the seller did the work.
11
Three Kinds of Answer
The gap between them is the product
MathSolverNow
TiersEffectSource
Rank the three
Point four nine,
point three two,
point zero five.
A meta-analysis of item-based feedback separated what the feedback actually said. An explanation beat the correct answer, and the correct answer beat being told right or wrong. Those three numbers are the specification for anything built here.
META-ANALYSIS OF ITEM-BASED FEEDBACK · 40 STUDIES, 70 EFFECT SIZESEFFECT SIZE
ELABORATED0.49AN EXPLANATION OF WHY
CORRECT RESPONSE0.32THE RIGHT ANSWER, GIVEN
VERIFICATION0.05RIGHT OR WRONG, NOTHING MORE
Reported range across the seventy effect sizes: -0.78 to 2.29. A meta-analysis is a distribution, not a promise.
PUBLISHED IN A PEER-REVIEWED EDUCATION JOURNAL, 2015BARS ARE TO SCALE
0.49elaborated
0.32correct answer
0.05right or wrong
How they position itThis is the clearest published specification for what a solver should return. A brand that builds to it can put the number in its own marketing.
12
Larger in Mathematics
The authors compared subjects, and said which one led
MathSolverNow
SubjectsQuoteCaveat
Read the sentence
The subject
where it mattered most.
The same meta-analysis compared effect sizes across subject areas. Its own summary sentence names mathematics ahead of social sciences, science and languages. That is a subject-level finding, and it is the one this address sits on.
FROM THE PUBLISHED ABSTRACTVERBATIM
“Larger effect sizes were found for mathematics compared with social sciences, science, and languages.”
MATHEMATICSSOCIAL SCIENCESSCIENCELANGUAGES
WHAT THE SENTENCE DOES NOT SAYIt does not give a per-subject number, and this page does not invent one. The ordering is the finding.
QUOTED, NOT PARAPHRASEDORDERING ONLY
Mathematicsnamed first
3subjects compared against
2015published
How they position itEvery adjacent subject could claim this literature. Only one of them is named first in the sentence, and that is worth an address.
13
The Hint Trap
Step-by-step help behaved differently from strategic help
MathSolverNow
FindingItemsDesign
See the split
Clicking through
until the answer appeared.
A 2025 study of student teachers in an interactive math environment measured help abuse directly. It predicted lower post-test scores and it mediated most of the negative effect of help-seeking. The effect showed up for step-by-step hints and not for strategic help.
INTERACTIVE LEARNING ENVIRONMENT · NUMBER CONVERSIONN = 322
Self-report item: “I playfully clicked through the hints until the correct answer I was looking for was given.”
STEP-BY-STEP HINTSNEGATIVEASSOCIATED WITH LOWER PERFORMANCE
STRATEGIC HELPNOT OBSERVEDNO SUCH ASSOCIATION REPORTED
AUTHORS’ OWN RECOMMENDATIONSTEP-BY-STEP HINTS “COULD BE FADED OUT”
THE RISK IS IN THE FORMAT OF THE HELPNOT IN ITS SPEED
-0.452help abuse on post-test
57%of the effect mediated
322participants
How they position itNaming the failure mode of your own category is how a serious operator opens a conversation with a district. The name gets you the meeting; this gets you the second one.
14
When It Is Taken Away
A field experiment, and what happened at the exam
MathSolverNow
DesignResultStatus
See both arms
Better with it.
Worse without it.
About a thousand students in grades nine through eleven used two versions of an assistant on math practice. Both raised performance while the tool was in hand. When access was removed for the exam, only one arm was worse off than students who never had it.
PRE-REGISTERED FIELD EXPERIMENT · GRADES 9-11 MATHWORKING PAPER
ARMWITH THE TOOL, ON PRACTICEON THE EXAM, TOOL REMOVED
Unrestricted+48%-17%
Guided, safeguarded+127%No significant effect
No accessbaselinebaseline
The practice figures are performance while the tool was in hand, which is the point of the study, not a learning gain.
NOT PEER REVIEWED AS OF SEPTEMBER 2026, AND A PUBLISHED DESIGN CRITIQUE EXISTSWE SAY SO
+127%practice, guided version
-17%exam, unrestricted arm
~1,000students
How they position itThe difference between the two arms is a product roadmap. A buyer who reads this card knows the seller understands what they are actually building.
15
What a Tutoring System Is Worth
Two meta-analyses, and the comparison that matters
MathSolverNow
EffectsAgainstCaveat
Read both
Against a classroom.
Against a person.
Computer tutoring systems have been measured against several baselines. They beat regular classroom instruction and non-adaptive software. Measured against a human tutor, the difference was not statistically significant.
INTELLIGENT TUTORING SYSTEMS, TWO META-ANALYSESEFFECT SIZE
VS REGULAR CLASSROOMg = 0.37CI .21 TO .53
VS NON-ADAPTIVE SOFTWAREg = 0.35p = .004
VS A HUMAN TUTORg = -0.25p = .302, NOT SIG.
THE OTHER META-ANALYSISFifty controlled evaluations, median effect 0.66, which its authors tie closely to how well the outcome test matched the instruction.
A NULL AGAINST HUMAN TUTORING IS NOT A WIN OVER ITWE PRINT THE P VALUE
0.66median, 50 evaluations
0.37against the classroom
Not sig.against human tutoring
How they position itA category this well studied rewards a brand that cites it accurately. The address is short enough to sit in a footnote and be remembered.
16
Timing Is Not the Claim
The finding that argues against this page, printed on it
MathSolverNow
ResultScopeReading
Read the null
g equals
zero point zero three.
A meta-analysis published in January 2026 pooled 51 studies to test immediate against delayed feedback. It found no significant difference on average. Any page that sells this address as faster learning is selling something the evidence does not support.
FEEDBACK TIMING IN COMPUTER-ASSISTED LEARNING · PUBLISHED JANUARY 2026AGAINST US
-0.300+0.30
POINT ESTIMATEg = 0.03
95% INTERVAL-0.08 to 0.13
THE INTERVAL CONTAINS ZEROp = .61
“Feedback timing does not significantly influence learning outcomes on average,” though its impact is “context-dependent.”
160 EFFECT SIZES, ROBUST VARIANCE ESTIMATIONSPEED IS NOT THE PRODUCT
0.03g, immediate vs delayed
51studies
p = .61not significant
How they position itThis card is why the page is credible. The name buys the moment; the buyer has to be worth the moment, and saying so out loud is the strongest thing on here.
17
What the Late Hour Costs
Fourteen days, five hundred and thirty-five students
MathSolverNow
DiarySleepNext day
Read the study
Trouble the
following day.
A daily-diary study of public high school students found sleep falling year over year while nightly study held roughly steady. On nights a student gave up sleep to study more than usual, the next day was measurably harder.
14-DAY DAILY DIARY · PUBLIC HIGH SCHOOL STUDENTSPEER REVIEWED, 2013
GRADESLEEP, HOURSSTUDY, HOURSSLEEP, TO SCALE
97.631.12
107.421.10
126.941.06
“If that student sacrifices sleep to study more than usual, he or she will have more trouble understanding material taught in class and be more likely to struggle on an assignment or test the following day.”
THE FINDING IS ABOUT TRADING SLEEP FOR STUDYNOT ABOUT HOMEWORK ITSELF
7.63 to 6.94hours of sleep, grade 9 to 12
~1 hournightly study
535students
How they position itA wellbeing claim that cites a real diary study, and states its limit in the same breath, survives contact with a school board.
18
What Cuts the Other Way
A government-data paper that disagrees with the easy version
MathSolverNow
SampleFindingUse
Read the counter
Homework was not
the thing.
A study of more than ten thousand adolescents using federal time-use data looked at what preceded later sleep. Passive leisure, public places and time with friends were associated with later sleep onset. Homework was not.
NATIONALLY REPRESENTATIVE TIME-USE SAMPLE · 2023AGAINST THE EASY VERSION
ASSOCIATED WITH LATER SLEEP ONSETPASSIVE LEISURE
ASSOCIATED WITH LATER SLEEP ONSETTIME IN PUBLIC LOCATIONS
ASSOCIATED WITH LATER SLEEP ONSETTIME WITH FRIENDS
NOT ASSOCIATEDHOMEWORK
NOT ASSOCIATEDACTIVE LEISURE
So this page never claims homework is keeping anybody up. It claims the help is closed, which is a different sentence and a checkable one.
A BUYER WILL FIND THIS PAPERIT IS ALREADY ON THE PAGE
10,341adolescents
15-18ages
Not homeworkwhat moved bedtime
How they position itPublishing the paper that refutes the lazy version of your pitch is the cheapest credibility available, and almost nobody in this category does it.
19
What the Register Shows
Two disclaimers, one cancellation, and an active opposer
MathSolverNow
RecordPatternRisk
Read the record
Disclaimed,
disclaimed, canceled.
The federal trademark record for this vocabulary is public and specific. It is not a clearance opinion and this page does not offer one. A buyer should have counsel review both the trademark position and any regulatory claim independently.
UNITED STATES TRADEMARK RECORD, AS PUBLISHEDNOT LEGAL ADVICE
MATHSOLVERABANDONED 1999FOR FAILURE TO RESPOND, NOT A REFUSAL ON THE MERITS
MATH PROBLEM SOLVERCANCELED 2012REGISTERED ONLY ON A SHOWING OF ACQUIRED DISTINCTIVENESS, MATH DISCLAIMED
TK SOLVERSOLVER DISCLAIMEDNO EXCLUSIVE RIGHT TO THE WORD, APART FROM THE MARK AS SHOWN
THE PATTERN IN TERM-PLUS-NOW REGISTRATIONSAcross the education and services records pulled, it is the other word that carries the disclaimer, not NOW.
WHAT A BUYER SHOULD KNOWA LARGE SOFTWARE REGISTRANT ACTIVELY OPPOSES NOW-FORMATIVE MARKS IN SOFTWARE CLASSES
ONE OPPOSED APPLICATION REGISTERED ANYWAY; ANOTHER WAS TERMINATEDHAVE COUNSEL REVIEW IT
MATHdisclaimed in a registration
SOLVERdisclaimed in a registration
2012the cancellation
How they position itHanding a buyer the risk alongside the record is what a seller does when the record is genuinely good. Anything less invites them to find it alone.
20
What Is Not Published
The holes, named, before a buyer finds them
MathSolverNow
GapsMethodHonest
See the gaps
Eight things we
will not print.
No federal source publishes the share of students doing homework at a given hour. No study measures when a student gives up. The evidence on timing is null. A buyer finds all of that in ten minutes, so it goes here instead.
WHAT DOES NOT EXIST, OR CUTS THE OTHER WAYEIGHT
Any federal figure for the share of students doing homework at a given hour
Any study measuring when in the evening a student gives up
Any significant effect of feedback timing on learning, across 51 studies
Any published average end time for the public high school day
Any homework figure by age newer than a 2008 visual essay
Any state agency running a homework hotline with published hours
Any evening spike large enough to survive three years of the same table
Any assurance about NOW-formative marks, which a large registrant opposes
WHAT WE DO INSTEADNAME THE SOURCE AND DATE IT
0federal figures on students by hour
0studies of the give-up moment
8gaps we can name
How they position itEvery page in this category oversells a moving target. The one that names its own holes is the one a serious buyer finishes reading.