Community Transit Ridership Jumped 24% After Lynnwood Link Restructure

30 Replies to “Community Transit Ridership Jumped 24% After Lynnwood Link Restructure”

1. Whenever you switch from direct express service to a trunk/feeder approach with feeder service and a transfer to rail, you increase the number of boardings, because some of the riders use two (unlinked) transit trips to make the trip they made before on one. To understand the ridership effect it would be helpful to know how the number of linked trips (a single rider’s trip regardless of transfers) has changed, which would give a better picture of whether there are more riders or just more boardings from transferring to rail. I’m guessing there’s some of both going on here, since feeder bus restructuring and SWIFT additions create more direct connections within Snohomish County, but without looking at linked trips it’s hard to know exactly what’s going on.

   1. In this case the trunk portion is a different agency, so it’s only one CT boarding the same as before.

      1. Yeah, that is what makes the ridership numbers so impressive. This does not include ridership at Link. If everyone just switched from riding a bus to riding a bus/train combination then ridership (on the bus) should be identical. But you also have riders who just ride the train. Thus Link poached some of the ridership from CT and yet CT ridership still went up.

         The change in ST Express service probably influenced the numbers. Before Lynnwood Link the only bus from Ash Way to Link was the 512 (which went to Northgate by then). Now those riders can take the 201 or 202 for that trip. The same thing is true for Everett. But it is likely that had only a minor impact on these numbers.

         I would say the only thing that could be misleading is that overall ridership for Community Transit was higher before the pandemic. The high point was the second quarter of 2019 when ridership reached an average of 36,800 riders a day. Ridership was also higher back in late 2008 and early 2009 when CT carried over 35,000 riders a day.

         Some of the ridership increase is likely due to the recovery from the pandemic. Some ridership has been lost to working from home but a lot of that is handled by Link. In other words the switch from peak to local service could not have come at a better time for the agency. Overall ridership for the agency is not as high as it once was but if you include Link ridership then overall Snohomish County fixed route transit ridership is probably higher than ever (although just barely).

         *The easiest way to calculate Snohomish County Link ridership is to look at ridership for the two stations and then double it. It needs to be doubled since the trip back (from Seattle) wouldn’t be counted otherwise. (Very few riders take Link between Mountlake Terrace and Lynnwood).

   2. It’s certainly true that the different service approach increases the boardings from one to two just for making the same trip.

      What I see here though is a shift from longer CT routes that run less frequently to shorter CT routes that run more frequently. Rather than have one rider on a bus for 50 minutes, it’s possible to have 2 riders for 25 minutes each.

      Along with that, didn’t some CT routes become more frequent? Applying the service hours to offer more frequent local service when it has been less frequent would seem to attract more riders.

      The situation reminds me of how the recent higher frequency of Link in North Seattle may have added riders to Link between that and Downtown Seattle. It’s tough to determine how much of Link’s ridership surge is due to crosslake opening versus the higher frequency within Seattle — but I think the higher frequency has influenced the ridership growth.

2. Thanks for highlighting Community Transit route 280, for years I have ridden the 280. It feels like Route 280 could almost double its frequency from hourly to every half-hour if CT would make the route less circuitous and more direct. The 280 route performs two circles both east and west of Lake Stevens through congested areas that make a half hour trip from Everett Station to Granite Falls require over an hour. It always feels silly riding a bus that is going in a circle, usually picking up or dropping off no passengers while performing these circles. My guess is a more direct route would significantly increase ridership numbers.

   With existing homes and new developments slated south of Granite Falls, an option to travel to the Lake Bosworth sub-area is worth reviewing. Hope CT will survey this route, solicit community feedback, and perhaps pilot a route extension.

   1. Yeah, I think there is a good case for changing the route. At first glance it is easy to assume that they get most of the ridership from Lake Stevens and Granite Falls is just an extra (coverage-oriented) bonus. Sure, it takes a while but if there aren’t that many riders then it is a reasonable approach. But if you look at the stop data (by going to this page and then choosing the route) you can see plenty of riders from Granite Falls. If anything, the tail is wagging the dog here. Other than the Lake Stevens Transit Center (which isn’t much of a detour) there isn’t much ridership in Lake Stevens. If you were only interested in ridership then you might just skip all of Lake Stevens other than the park and ride. I certainly wouldn’t do that. The first issue is cost. Granite Falls is farther away and thus more expensive to serve. But even beyond that it is worth serving some of the areas in Lake Stevens even if ridership is relatively low. The main takeaway is what you wrote — it could certainly be improved by avoiding the loops. Just a straight-forward turn on 20th & Grade seems like a nice improvement. Very few riders would be hurt by this and Lake Stevens has microtransit service.

      Complicating things even further is the new development along SR-9 like the new apartment building going up close to Soper Hill Road. The only bus service in the area is the 209 which goes between the Lake Stevens Transit Center and Smoky Point. If riders in that apartment want a direct bus to Everett they have a long walk along the highway or an even long walk on Soper Hill Road (which doesn’t have sidewalks). This is another argument for just following the two highways and just adding a couple bus stops along the way. Another option would be to add another roundabout at Lake Drive and SR-92. That way the bus could detour right by the development there as well as the elementary school before heading towards Granite Falls. All that being said it might just be robbing Peter to pay Paul. The area lacks a good grid (because of natural features like the lake) which make it a lot more difficult to serve.

      I could definitely see the route being modified although I’m not sure the ideal approach. It could use exactly what you suggested. Some analysis and discussion along with some changes.

      1. SR99 or SR 9?

         1. Sorry, I meant State Route 9. I’m so used to typing “State Route 99” that I accidentally wrote that. I’ll correct it.

3. Did anyone notice the double decker on the 271 this morning? This makes me think CT might be exploring more opportunities with them, or it could be just a vehicle selection choice. This makes me think of how the 424 had one trip in the morning that could sometimes run an ADL bus, but now the 908 doesnt.

   1. CT will keep running the double-decker buses until they reach end-of-life. I think they have a few more years left on them so we’ll be seeing them around until all the double talls are retired.

      1. Why is CT ending the double tall’s? Does ADL no longer manufacture to the US?

         1. I assume it’s harder/more expensive to maintain such a fleet and they want to pare down to just artics and 40′ buses rather than three different fleets.

         2. They cost more to run and the extra capacity isn’t needed. The main benefit of double-talls was the higher capacity in the same footprint as a ~40-foot bus, which was important when CT was parking commuter buses in/near downtown Seattle.

         3. See the 2nd and 3rd paragraphs, here: https://seattletransitblog.com/2026/06/13/community-transit-ran-its-last-bus-to-seattle/

         4. Will their artics be phased out as well (except on Swift) once they retire them? As a recent new rider and resident of Snohomish County, seems like the only time those artics get full are when they are used on the 510. Seems to me CT should continue with the more frequent local and smaller buses. Just an observation.

   2. CT official said in a board meeting last year that service change they did to 424 in Monroe area was to prepare running double tall on future 908.
      I thought that meant they would primarily use ADL for 908, but they didn’t.

   3. The 280 could also use them… I just wish they could extend it to Lynnwood and delete the 903.

4. Does CT provide rides attracted per platform hour? The hours on most routes changed. Is the network more efficient? I expect so, but this data does not demonstrate it.

   1. See the next comment. All of the data here was provided after Micheal requested it. I believe they provided a huge spreadsheet with stop data. He then took that data and was able to create the reports on the other page as well as the summaries per route found on this page. It is possible they would provide rides-per-service-hour (and similar metrics) but it is also quite possible they never calculated it. They are a pretty small agency. They haven’t been particularly ridership-oriented although it is quite possible they do this type of work when analyzing Swift routes.

5. This is an impressive article. I’ve wondered about the ridership of Community Transit (CT) routes for a long time. Unlike other local agencies, CT doesn’t publish ridership by route except for the occasional press release about a Swift route. So far as I know, this is the first time anyone has made this information easily available. To have the stop data on top of that is a huge bonus. Excellent work Michael.

   To get an idea of what CT provides, here are some links: This is the report from 2019. There is a lot of information in there but nothing about the ridership of individual routes. The only thing they have is on page 26 where they list ridership based on each route-type (Swift, Corridor-Based, Local Feeder, Suburban/Rural and Commuter). Other more recent reports (like this one) just show overall ridership.

   1. Great analysis, Michael, and great to see the ridership growth. I bet the work on I-5 also provided an incentive for riders to explore CT’s alternatives.

6. Thanks for the info, Michael.

   With new proof that the 111 is the least used route in the entirety of Community Transit, what are we going to do about it? It has a long history though:

   When Metro started in 1973, they ran a bus route numbered the 377 which ran between Brier and Northgate. It continued to Downtown Seattle during the peak periods.

   The 377 was truncated in Horizon View (soon extended to Lynnwood). The latter was replaced by one of Community Transit’s cream buses, the 477 which ran between Brier and Downtown Seattle. The cream buses were operated by Metro, but soon then fully operated by Community Transit. Rapid transit between Snohomish County and Seattle had been a vision ever since, starting in the 80’s with the cream buses.

   The 377 was replaced by the 347 in 2003, with no coverage by Metro from Lynnwood to MLT. Instead, the 112 ran on 44th Ave (which back in the day it ran from Lynnwood to Edmonds CC aka Edmonds College which was truncated in MLT but extended to Ash Way in 2012 as part of the big service cuts). The 347 ran between MLT and Northgate via Jackson Park.

   The 477 was deleted in 2011 along with the 408 (which ran between MLT and Downtown Seattle) when the MLT Freeway Station opened. As part of the new freeway station, Community Transit considered two options for a new local Brier route with the number 111:
   – One option ran between Lynnwood and Brier via 48th Ave W.
   – The second option ran between MLT and Brier which is what we know today.

   The 347 was replaced with the 333, 365, and increased 348 service when Lynnwood Link opened.

   So that’s how the 111 started, it had more trips back when it started in 2011 which were chopped first in 2012 then in 2020 until it just had 2 round-trips per weekday.

   1.25 riders a trip!? That’s abysmal. Brier already has public transit problems given that it lost Metro FLEX two months ago and Community Transit is doing NOTHING about it. What will Brier do? Will it push for improved buses and a possible rezone or are we just going to keep this in perpetuity?

7. They should merge ST 512 + CT 905 and make it 1 efficient long distance express route. Frequency and availability of the 905 is pretty awful.

   1. “They should merge ST 512 + CT 905 and make it 1 efficient long distance express route.”

      And I thought I was the only one who thought that… Oh! Also merge it with the freeway segment on the 201/202.

      “Frequency and availability of the 905 is pretty awful.”

      I think everyone agrees.

8. I cannot drive and I havd to rely on lite rail and busses to f see t places. Since light rail has started I love we how quickly I can get from Mt lk terrace to Seattle but I now have way less options when I gavd yo get to ballard or Fremont for dr visits. I used to be able to transfer my way but now I have to pay Uber or lyft to get my from light rail to the final destination which js very speedy. I know you a rd not metro but I wish that bus systems could have a platform like lyft that could help you get from I m light rail platforms to your needed destination without working over your wallet. Just a idea.

   1. But according to everyone on here, we should close bus stops to funnel into already high ridership routes with already decent enough frequency! “low” ridership is a reason to take away coverage from the people who need it most.

      1. @SKR

         You have it entirely backwards. High ridership local routes serve many, many more riders that need it than low ridership coverage routes. Coverage routes generally provide service to far fewer people.

         High ridership local routes also serve people that travel into the city, since even once you arrive, that doesn’t matter if you can’t get where you want to go. Gina’s example is illustrative: Link-Fremont and Link-Ballard should both be very high frequency local routes, not a potential long wait for a transfer.

         The biggest gap for Metro is service hours. Every city should have (per 50k-100k residents or so) a few frequent local routes, a handful of coverage routes, and an express (or at least a reasonable connection) to Seattle and perhaps Bellevue. But Metro doesn’t have enough service hours to provide that across all of King County.

      2. But according to everyone on here, we should close bus stops to funnel into already high ridership routes with already decent enough frequency!

         So far as I know, no one — I mean literally no one — has proposed we get rid of all of the coverage routes. Thus the idea that everyone wants this is absurd.

         Most people are well aware of the ridership/coverage trade-off (https://humantransit.org/2018/02/basics-the-ridership-coverage-tradeoff.html). Some think we should spend a little more money on coverage, others on ridership. But what people mostly want is a system that is better on both counts. For example, imagine two bus routes running north-south a block away from each other. Each runs every half hour. They are the only bus routes in the area. Ridership is bad because the buses only run every half hour. If you combined them you could run every fifteen minutes and you would get a lot more riders. But it is also bad from a coverage standpoint. You haven’t added much coverage at all — just one block to one side.

         This is an extreme example but there are many other cases that have the same basic problem. Routing neither provides good ridership nor good coverage. But while some cases (like the fictional one I mentioned) are easy flaws to notice, most involve tricky trade-offs. Ridership increases with frequency but it also increases when you avoid a transfer. These are contradictory ideas when it comes to planning. Speed matters but so does the distance to a bus stop. Again, this is contradictory. That is why transit planning is full of subtle trade-offs. This includes the fundamental trade-off between ridership and coverage.

   2. I’m not sure what part of Fremont you are trying to get to. The 31/32 go from the U-District Station to 35th and then Fremont. The 62 goes from Roosevelt Station to Stone Way and then Fremont. Thus they serve central and eastern Fremont fairly well. Western Fremont requires a longer walk or transfer (to the 40) while the north end of Fremont is only served by the 5 (which doesn’t connect well to these buses and only connects to Link downtown).

      The 31/32 and 62 are both good buses with decent all-day frequency (fifteen minutes midday). The only drawback is that they are relatively slow. While the 62 could be improved, both corridors will never be particular fast. There are just too many busy cross streets and too many turns. It is a typical urban environment with no expressway connecting neighborhoods. As jd mentioned, that is why a subway line to Ballard (which would include Fremont) would make such a big difference. It would faster than driving — at noon. But instead Sound Transit is focused on corridors that are already fairly fast (like West Seattle to downtown).

      So yeah, while I’m generally highly critical of Metro I don’t think the routing is the problem here (other than the aforementioned flaw with the 62 which is actually the fault of SDOT). The problem is that it s a fundamentally slow surface connection that should be replaced by a subway line but ST has other, misguided priorities.

   3. What did you lose? There was never an express bus from Mountlake Terrace to Ballard or Fremont. The express buses to the U-District stopped three blocks closer at I-5 & 45th compared to U-District Station, but that’s just a 1 or 2 minute difference. In both cases you’d transfer to the 44 to Ballard. The 44 has fluctuated between 10- and 15-minute frequency depending on Metro’s finances and driver availability, but that has nothing to do with Link or the Metro restructure, and Sound Transit and Community Transit (the owners of the former express buses) have no control over it. To get to Fremont, you could take the 44 and walk, or if you want to take the 31/32, the Link transfer is closer than freeway buses stopping at the 45th stop. So what have you lost?

      What you’ve gained is not getting stuck in freeway congestion, or lane closures due to car collisions. If those didn’t affect you personally, they affected others who took I-5 express buses.

Comments are closed.